본문으로 건너뛰기

#Review

3409개의 포스트

[논문리뷰] Watch Before You Answer: Learning from Visually Grounded Post-Training

본 논문은 오직 Visually Grounded 질문만을 사용하여 모델을 post-training하는 VidGround 프레임워크를 제안한다. 저자들은 GPT-5-mini와 같은 강력한 언어 모델을 사용하여 텍스트만으로 답변 가능한 TA 질문을 필터링하고, 나머지 VG 질문만을 학습 데이터로 선별하였다.
#Review#Vision-Language Models#Video Understanding#Post-Training#Linguistic Bias#Reinforcement Learning#Data Curation#Visually Grounded Reasoning
댓글 수 로딩 중

[논문리뷰] Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding

본 논문은 12명의 전문 주석자와 50명의 검토자가 3,300시간을 투입하여 구축한 Video-MME-v2 데이터셋을 통해 모델의 역량을 평가한다. 제안하는 방법론은 비디오 이해 능력을 3단계로 체계화하고, 질문을 그룹화하여 Consistency와 Coherence를 검증하는 그룹 기반 평가 전략을 포함한다.
#Review#Video Understanding#Multimodal Large Language Models#Benchmark#Reasoning Coherence#Capability Consistency#Evaluation Hierarchy#Non-linear Scoring
댓글 수 로딩 중

[논문리뷰] Vanast: Virtual Try-On with Human Image Animation via Synthetic Triplet Supervision

본 논문은 Vanast를 통해 단일 단계에서 고충실도의 애니메이션을 합성하는 end-to-end 시스템을 제안합니다. 핵심 방법론인 Dual Module은 인간 애니메이션 모듈(HAM)과 의류 전송 모듈(GTM)로 나뉘어 사전 학습된 Video DiT 백본을 공유하며, 학습 시에는 이 두 모듈만을 최적화하여 빠른 수렴을 도모합니다 .
#Review#Virtual Try-On#Human Image Animation#Diffusion Transformer#Synthetic Triplet Supervision#Dual Module Architecture#Zero-shot Garment Interpolation
댓글 수 로딩 중

[논문리뷰] ThinkTwice: Jointly Optimizing Large Language Models for Reasoning and Self-Refinement

본 논문은 Reasoning 최적화와 Self-Refinement 최적화를 하나의 GRPO 프레임워크 안에서 결합한 ThinkTwice를 제안합니다. ThinkTwice는 각 훈련 단계에서 모델이 먼저 Reasoning 문제를 풀고, 동일한 문제에 대해 자신의 이전 답변을 개선하는(Thinking twice) 과정을 연속적으로 수행합니다 .
#Review#Large Language Models#Reinforcement Learning#Reasoning#Self-Refinement#RLVR#Policy Optimization#Implicit Curriculum
댓글 수 로딩 중

[논문리뷰] Paper Circle: An Open-source Multi-agent Research Discovery and Analysis Framework

본 논문은 문헌 탐색부터 분석, 비평까지 연구 전 과정을 지원하는 다중 에이전트 플랫폼인 Paper Circle을 제안한다. 시스템은 Discovery Pipeline과 Analysis Pipeline으로 구성되며, 연구자는 CodeAgent(CoA) 기반의 오케스트레이션 프레임워크를 통해 결정론적이고 추적 가능한 결과를 얻을 수 있다 , .
#Review#Multi-agent Systems#Research Discovery#Knowledge Graph#Large Language Models#Academic Literature Review
댓글 수 로딩 중

[논문리뷰] MegaTrain: Full Precision Training of 100B+ Parameter Large Language Models on a Single GPU

MegaTrain은 파라미터와 옵티마이저 상태를 호스트 메모리에 영구 저장하고, GPU는 레이어별 연산 시에만 파라미터를 스트리밍하는 구조를 채택한다. 저자들은 CPU-GPU 대역폭 병목을 극복하기 위해 Pipelined Double-Buffered Execution Engine을 도입하여 연산과 데이터 이동을 비동기적으로 중첩시켰다 .
#Review#Large Language Models#Memory-Centric#CPU-GPU Offloading#Parameter Streaming#Stateless Execution#Pipelined Execution#Single-GPU Training
댓글 수 로딩 중

[논문리뷰] MedGemma 1.5 Technical Report

본 논문은 Gemma3 아키텍처를 기반으로 MedSigLIP Vision Encoder를 결합하고, 고차원 의료 데이터를 위한 특화 전처리 파이프라인을 도입한다. 3D 영상의 경우 axial 슬라이드를 시퀀스로 변환하여 인코딩하며, WSI는tissue mask 기반의 확률적 패치 샘플링을 통해 메모리 효율을 높였다 .
#Review#MedGemma#Multimodal Learning#Medical Imaging#Foundation Models#Volumetric Analysis#Histopathology#Clinical Reasoning
댓글 수 로딩 중

[논문리뷰] Experience Transfer for Multimodal LLM Agents in Minecraft Game

본 논문은 Echo 프레임워크를 통해 환경 지식을 5가지 전이 차원으로 분해하고 CSD를 통해 이를 통일된 의미론적 형태로 변환하여 관리합니다 . CSD는 시각적 및 텍스트 정보를 벡터화된 임베딩과 결합하여 메모리 뱅크에 저장하며, 이를 통해 ICAL 알고리즘이 관련 경험을 정밀하게 검색할 수 있도록 지원합니다 .
#Review#Multimodal LLM Agent#Experience Transfer#In-Context Analogy Learning (ICAL)#Minecraft#Contextual State Descriptor (CSD)#Embodied Intelligence
댓글 수 로딩 중

[논문리뷰] DARE: Diffusion Large Language Models Alignment and Reinforcement Executor

본 논문은 dLLM을 위한 통합 후학습 및 평가 프레임워크인 DARE (dLLMs Alignment and Reinforcement Executor)를 제안한다. DARE는 verl과 OpenCompass를 기반으로 하며, 다양한 모델 패밀리(MDLM 및 BDLM)를 동일한 실행 스택에서 처리할 수 있도록 설계되었다 .
#Review#Diffusion Large Language Models#Post-Training#Reinforcement Learning#Unified Framework#Systems Optimization
댓글 수 로딩 중

[논문리뷰] Can Natural Image Autoencoders Compactly Tokenize fMRI Volumes for Long-Range Dynamics Modeling?

저자들은 자연 영상에 사전 학습된 2D DCAE를 사용하여 3D fMRI 볼륨을 27개의 컴팩트한 토큰으로 효율적으로 압축하는 토큰화 방식을 제안합니다 . 이 토큰화된 데이터는 메모리 효율적인 Transformer 아키텍처에 입력되며, MTM을 통해 자기지도 사전 학습을 수행함으로써 하류 작업(downstream tasks) 성능을 향상합니다.
#Review#fMRI#Transformer#Tokenization#Deep Compression Autoencoder#Self-supervised Learning#Long-range Dynamics
댓글 수 로딩 중

[논문리뷰] Beyond Accuracy: Unveiling Inefficiency Patterns in Tool-Integrated Reasoning

본 연구는 transformer 추론의 제1 원리에 기반하여 prefill과 decode 단계의 비대칭적 비용을 모델링한 하드웨어 인식 지표인 PTE를 제안한다. PTE는 memory-bound인 decode 비용을 compute-bound인 prefill 토큰 단위로 환산하여 통합함으로써, 하드웨어 프로필에 무관한 일관된 효율성 평가를 가능하게 한다.
#Review#Tool-Integrated Reasoning (TIR)#KV-Cache#PTE (Prefill Token Equivalents)#Inference Efficiency#Hardware-Aware Metric#LLM Agent
댓글 수 로딩 중

[논문리뷰] Action Images: End-to-End Policy Learning via Multiview Video Generation

본 논문은 7-DoF 로봇 제어 입력을 Action Images로 변환하여 시각적 표현으로 통합하는 방식을 취합니다 . 제안 모델은 Wan 2.2 비디오 백본을 기반으로 하며, RGB 비디오와 Action Images를 결합하여 비디오 공간에서 물리적 동역학을 모델링합니다 .
#Review#World Action Model#Robot Policy Learning#Multiview Video Generation#Pixel-grounded Representation#Zero-shot Policy
댓글 수 로딩 중

[논문리뷰] Vero: An Open RL Recipe for General Visual Reasoning

저자들은 6개 범주를 아우르는 600K 샘플의 Vero-600K를 구축하고, 태스크별로 세분화된 보상 함수를 적용하는 GSPO 기반의 단일 단계 RL 학습을 수행한다 . 데이터 정제 과정에서 모델 기반의 문항 필터링과 정답 정규화를 통해 학습 품질을 극대화하였다.
#Review#Vision-Language Models#Reinforcement Learning#Visual Reasoning#Multi-task Learning#Chain-of-Thought#Data Diversity#Reward Engineering
댓글 수 로딩 중

[논문리뷰] Unifying Group-Relative and Self-Distillation Policy Optimization via Sample Routing

저자들은 샘플의 학습 상태에 따라 적절한 최적화 방식을 할당하는 SRPO (Sample-Routed Policy Optimization)를 제안합니다 . SRPO는 정답 샘플에 대해서는 GRPO의 보상 정렬(reward-aligned) 강화를 적용하고, 오류 샘플 중 피드백 정보가 가용한 경우에는 SDPO의 정밀한 logit 수준 교정을 적용합니다.
#Review#RLVR#GRPO#SDPO#Sample Routing#Policy Optimization#Self-Distillation
댓글 수 로딩 중

[논문리뷰] The Geometric Alignment Tax: Tokenization vs. Continuous Geometry in Scientific Foundation Models

본 연구는 합성 동역학 시스템을 활용한 제어된 실험을 통해 이산적 토큰화(tokenization)가 기하학적 불안정성의 원인임을 입증합니다. 동일한 인코더 백본을 유지한 채 출력 헤드만 이산적인 Cross-Entropy 대신 연속적인 MSE 기반으로 변경했을 때, 기하학적 왜곡이 최대 8.5배 감소함을 확인하였습니다 .
#Review#Geometric Alignment Tax#Foundation Models#Tokenization#Manifold Fracture#Rate-Distortion Theory
댓글 수 로딩 중

[논문리뷰] SpatialEdit: Benchmarking Fine-Grained Image Spatial Editing

본 논문은 공간 편집 능력을 확장하기 위해 체계적인 데이터 생성 엔진인 SpatialEdit-500k와 기하학적 정밀성을 평가하는 SpatialEdit-Bench를 제안한다. SpatialEdit-500k는 Blender를 사용하여 다양한 객체와 장면에서 카메라 궤적과 객체 변환을 포함한 50만 개의 쌍(paired) 데이터를 생성하여 학습을 지원한다 .
#Review#Image Spatial Editing#Benchmark#Dataset#Geometry-Aware Evaluation#Camera Manipulation#Object Manipulation#Multimodal Large Models
댓글 수 로딩 중

[논문리뷰] SkillX: Automatically Constructing Skill Knowledge Bases for Agents

저자들은 에이전트의 경험을 계층적으로 구조화하여 재사용성을 극대화하는 SkillX 프레임워크를 제안합니다 . SkillX는 고성능 backbone 에이전트인 GLM-4.6을 사용하여 훈련 태스크를 수행하고, 성공한 trajectory로부터 Planning, Functional, Atomic 단계의 skill을 증류(distill)합니다.
#Review#LLM Agents#Skill Knowledge Base#Hierarchical Representation#Autonomous Experience Learning#Agent Self-Evolution#Tool Use
댓글 수 로딩 중

[논문리뷰] MinerU2.5-Pro: Pushing the Limits of Data-Centric Document Parsing at Scale

본 논문은 데이터의 범위, 정보성, 정확도를 극대화하기 위해 DDAS, CMCV, Judge-and-Refine 파이프라인으로 구성된 데이터 엔진을 제안한다 . DDAS는 샘플링의 다양성과 난이도를 동시에 고려하여 학습 데이터를 10M 이하에서 65.5M으로 확장하였으며 , CMCV는 모델 간 불일치를 활용하여 데이터의 난이도를 등급화한다.
#Review#Document Parsing#Data-Centric AI#Vision-Language Model#Progressive Training#Data Engine
댓글 수 로딩 중

[논문리뷰] LightThinker++: From Reasoning Compression to Memory Management

저자들은 암시적 압축에서 시작하여 명시적 행동 수준의 관리로 진화하는 LightThinker 계열 모델을 제안한다. LightThinker는 gist tokens와 특수 설계된 attention mask를 활용하여 긴 사고 과정을 컴팩트한 표현으로 변환한다 .
#Review#Large Language Models#Reasoning Compression#Memory Management#Agentic Reasoning#Context Optimization
댓글 수 로딩 중

[논문리뷰] LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Models

저자들은 로봇 조작 지시문의 핵심 구성 요소인 행동과 객체를 기반으로 43개의 정밀한 변형 유형을 포함하는 LIBERO-Para를 구축하였다. 또한, 모델의 성공 여부뿐만 아니라 원문 지시문과 파라프레이즈 간의 키워드 유사도(SKS_K)와 구조적 유사도(STS_T)를 결합한 PRIDE 메트릭을 제안하여 보다 해석 가능한 견고성 평가를 수행한다 .
#Review#Vision-Language-Action (VLA) Models#Paraphrase Robustness#Robotic Manipulation#Diagnostic Benchmark#PRIDE Metric#Object Grounding#Trajectory Divergence
댓글 수 로딩 중

[논문리뷰] ClawArena: Benchmarking AI Agents in Evolving Information Environments

저자들은 에이전트의 성능을 다차원적으로 평가하기 위해 8개 전문 도메인, 64개 시나리오, 1,879개 라운드로 구성된 ClawArena 벤치마크를 구축하였습니다 . 각 시나리오는 숨겨진 Ground Truth를 바탕으로 구성되며, 에이전트는 노이즈가 섞인 부분적인 정보만을 관찰하여 추론해야 합니다.
#Review#AI Agents#Benchmark#Information Environments#Multi-source Reasoning#Belief Revision#Implicit Personalization
댓글 수 로딩 중

[논문리뷰] Xpertbench: Expert Level Tasks with Rubrics-Based Evaluation

본 논문은 1,000명 이상의 현업 전문가가 참여하여 구축한 1,346개의 전문 작업으로 구성된 XpertBench 프레임워크를 제안한다. 평가 신뢰성을 위해 각 작업은 15~40개의 가중치가 부여된 원자적 체크포인트 기반의 Rubrics를 따르며, 이를 평가하기 위해 ShotJudge 패러다임을 도입했다.
#Review#XpertBench#LLM Evaluation#Expert-level Cognition#Rubrics-based Assessment#ShotJudge#Ecological Validity
댓글 수 로딩 중

[논문리뷰] Token Warping Helps MLLMs Look from Nearby Viewpoints

본 논문은 토큰을 변환 단위로 사용하는 Token Warping 프레임워크를 제안하며, 특히 Backward Token Warping이 안정성과 의미론적 일관성 측면에서 가장 우수함을 입증한다. 와 는 MLLM 토큰이 위치 잡음에 강건하다는 점을 증명하며, 이를 바탕으로 시점 변환 시 토큰 기반의 역투영 기법을 적용한다.
#Review#Multimodal Large Language Models#Token Warping#Viewpoint-Aware Reasoning#Spatial Reasoning#Mental Imagery
댓글 수 로딩 중

[논문리뷰] AgentSocialBench: Evaluating Privacy Risks in Human-Centered Agentic Social Networks

저자들은 AgentSocialBench를 도입하여 7개 카테고리, 300개 이상의 시나리오를 통해 에이전트의 프라이버시 보존 성능을 체계적으로 평가합니다 . 제안된 프레임워크는 사용자 프로파일의 민감도 계층, 지향성 사회 그래프, 그리고 명시적 프라이버시 경계를 포함합니다.
#Review#LLM Agents#Agentic Social Networks#Privacy Preservation#Contextual Integrity#Information Abstraction#Abstraction Paradox
댓글 수 로딩 중

[논문리뷰] AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents

저자들은 위험 카테고리(Risk categories)와 공격 전략(Attack strategies)을 포함한 체계적인 분류법을 설계하고, 이를 바탕으로 2,653개의 유해한 작업 인스턴스를 구축하였습니다 . 제안된 AgentHazard는 에이전트가 샌드박스 환경 내에서 작업을 수행하게 한 뒤, 전체 실행 경로를 심사하여 유해성 여부를 판별합니다.
#Review#Computer-Use Agents#Agent Safety#Benchmark#Harmful Behavior#Trajectory-level Evaluation#Multi-step Reasoning
댓글 수 로딩 중

[논문리뷰] A Simple Baseline for Streaming Video Understanding

본 연구에서 제안하는 SimpleStream은 추가적인 학습이나 별도의 메모리 구조 없이, 오직 마지막 $N$개의 프레임을 오프-더-셸프 VLM에 전달하는 최소주의적 추론 정책을 채택합니다 . 저자들은 OVO-Bench와 StreamingBench를 통해 13개의 주요 온라인 및 오프라인 비디오 LLM 베이스라인과 성능을 비교하였습니다.
#Review#Streaming Video Understanding#Video LLM#Simple Baseline#Perception-Memory Trade-off#OVO-Bench
댓글 수 로딩 중

[논문리뷰] VOID: Video Object and Interaction Deletion

본 연구는 CogVideoX 확산 모델을 기반으로, 물리적 인과 관계를 반영하는 카운터팩추얼 생성 모델을 구축하였습니다. 먼저 Kubric과 HUMOTO를 통해 객체 제거 전후의 물리적 역학 변화를 학습하고, VLM을 활용해 영상 내 영향받는 영역을 실시간으로 추론하여 Quadmask를 생성함으로써 모델의 생성 범위를 명확히 제한합니다.
#Review#Video Object Removal#Counterfactual Reasoning#Video Diffusion Models#Interaction-Aware Masking#Vision-Language Models
댓글 수 로딩 중

[논문리뷰] Tex3D: Objects as Attack Surfaces via Adversarial 3D Textures for Vision-Language-Action Models

본 논문은 Tex3D를 제안하여 VLA 시뮬레이션 환경 내에서 adversarial 3D 텍스처를 end-to-end로 최적화합니다. 제안하는 FBD는 MuJoCo에서 배경을 렌더링하고 Nvdiffrast에서 객체를 렌더링하여 두 렌더러 간의 MVP(Model-View-Projection) 및 조명 파라미터를 동기화함으로써 미분 가능한 경로를 확보합니다 .
#Review#VLA Models#3D Adversarial Textures#Embodied Robustness#Differentiable Rendering#Foreground-Background Decoupling
댓글 수 로딩 중

[논문리뷰] T5Gemma-TTS Technical Report

본 논문은 T5Gemma 모델을 백본으로 활용하여 방대한 언어적 지식을 음성 생성 영역으로 전이합니다. 제안된 모델은 XCodec2를 사용하여 오디오를 토큰화하며, Cross-attention 레이어마다 PM-RoPE를 적용하여 생성 진행 상황을 실시간으로 감시함으로써 정밀한 duration control을 달성합니다 .
#Review#text-to-speech#zero-shot voice cloning#encoder-decoder#PM-RoPE#multilingual evaluation
댓글 수 로딩 중

[논문리뷰] Steerable Visual Representations

저자들은 텍스트 프롬프트로 ViT의 내부 레이어를 직접 제어하는 SteerViT를 제안합니다. SteerViT는 frozen된 ViT 블록들 사이에 21M 개의 파라미터만 추가하는 경량화된 cross-attention 레이어를 삽입하여 텍스트 정보를 주입합니다 .
#Review#Steerable Visual Representations#Vision Transformers#Early Fusion#Cross-Attention#Text-Conditioned Vision#Representational Quality#Zero-Shot Generalization
댓글 수 로딩 중

[논문리뷰] SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization

저자들은 Skill0 프레임워크를 제안하며, 이는 ICRL과 Dynamic Curriculum을 결합하여 스킬을 단계적으로 내재화한다 . 학습 초기에는 풍부한 스킬 문맥을 제공하여 에이전트의 초기 탐색을 돕고, 학습이 진행됨에 따라 스킬 예산을 선형적으로 감소시켜 에이전트가 스스로 최적의 전략을 내재화하도록 유도한다.
#Review#In-Context Reinforcement Learning#Skill Internalization#Agentic Agents#Dynamic Curriculum#Context Compression
댓글 수 로딩 중

[논문리뷰] Omni123: Exploring 3D Native Foundation Models with Limited 3D Data by Unifying Text to 2D and 3D Generation

Yihao Zhi이 arXiv에 게시한 'Omni123: Exploring 3D Native Foundation Models with Limited 3D Data by Unifying Text to 2D and 3D Generation' 논문에 대한 자세한 리뷰입니다.
#Review#3D Native Foundation Models#Autoregressive Framework#Cross-modal Generative Consistency#Interleaved Training#3D Data Scarcity#Instruction-based 3D Editing
댓글 수 로딩 중

[논문리뷰] LatentUM: Unleashing the Potential of Interleaved Cross-Modal Reasoning via a Latent-Space Unified Model

본 논문은 시각적 이해와 생성을 공유된 semantic latent space에서 통합하는 LatentUM을 제안한다. 핵심 방법론인 MBAQ는 VLM의 출력 분포를 보존하도록 설계되어, 시각적 특징을 복원 중심이 아닌 이해 중심의 디스크리트 토큰으로 양자화한다 .
#Review#Unified Models#Cross-Modal Reasoning#Semantic Latent Space#MBAQ#Mixture-of-Modal Experts
댓글 수 로딩 중

[논문리뷰] Gated Condition Injection without Multimodal Attention: Towards Controllable Linear-Attention Transformers

본 논문은 Linear Attention 기반 모델을 위한 통합 게이트 조건 주입 모듈인 GateControl을 제안합니다. 이 방식은 학습 가능한 게이트를 통해 토큰별로 중요한 조건 정보만을 선택적으로 보존함으로써, 기존의 Multimodal Attention 없이도 강력한 제어 성능을 달성합니다.
#Review#Diffusion Transformer#Linear Attention#Controllable Generation#Gated Condition Injection#On-device AI
댓글 수 로딩 중

[논문리뷰] Friends and Grandmothers in Silico: Localizing Entity Cells in Language Models

본 논문은 templated prompts를 사용하여 특정 개체에 반응하는 뉴런을 추출하고, 이를 인과적 개입(Causal Intervention)을 통해 검증하는 파이프라인을 제안합니다. 먼저, 여러 프롬프트에서 안정적으로 활성화되는 뉴런을 순위화하여 Entity Cells를 식별합니다.
#Review#Mechanistic Interpretability#LLM#Entity Cells#Factual Recall#Causal Intervention#MLP Neurons#Canonicalization
댓글 수 로딩 중

[논문리뷰] DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models

본 논문은 LLaMA-Factory의 모델 관리 및 최적화 기능을 보존하면서, 데이터 최적화 전략을 모듈형으로 통합한 DataFlex를 제안합니다. DataFlex는 7개의 데이터 선택, 2개의 데이터 혼합, 1개의 재가중치 알고리즘을 통합하여 단일 인터페이스 하에서 연구 및 실행할 수 있도록 설계되었습니다 .
#Review#Data-Centric AI#Large Language Models#Dynamic Training#Data Selection#Data Mixture#Data Reweighting#LLaMA-Factory
댓글 수 로딩 중

[논문리뷰] Automatic Image-Level Morphological Trait Annotation for Organismal Images

본 논문은 Sparse Autoencoders(SAE)와 Multimodal Large Language Models(MLLM)을 결합한 모듈형 자동 주석 파이프라인을 제안합니다. 우선 DINOv2 백본을 통해 추출된 특징에 SAE를 학습시켜 공간적으로 명확한 형태학적 부분을 담당하는 뉴런을 식별합니다.
#Review#Sparse Autoencoders#Morphological Trait Annotation#Multimodal Large Language Models#Fine-grained Visual Recognition#Biological Foundation Models
댓글 수 로딩 중

[논문리뷰] AIBench: Evaluating Visual-Logical Consistency in Academic Illustration Generation

본 논문은 학술 도해의 논리적 정확성과 미학적 품질을 분리하여 평가하는 AIBench를 제안한다. 논리 평가를 위해 논문 본문에서 논리 그래프를 추출하고, 이를 기반으로 4단계(Component, Topology, Phase, Semantics) 수준의 VQA 질문 세트를 자동으로 생성 및 인간 전문가가 검수한 데이터를 구축하였다 .
#Review#Academic Illustration#Visual-Logical Consistency#Benchmark#VQA#Test-Time Scaling#Multimodal Evaluation
댓글 수 로딩 중

[논문리뷰] ViGoR-Bench: How Far Are Visual Generative Models From Zero-Shot Visual Reasoners?

본 논문은 시각적 추론을 측정하기 위해 Physical, Knowledge, Symbolic 세 가지 영역을 포괄하는 ViGoR-Bench를 제안합니다 . 이 프레임워크는 Ground Truth(GT)를 기반으로 하여 모델의 생성 과정을 정밀하게 감시하는 Dual-Track Process-Outcome Evaluation 프로토콜을 구현합니다 .
#Review#Visual Generative Models#Visual Reasoning#Evaluation Benchmark#Chain-of-Thought#Process-Outcome Evaluation
댓글 수 로딩 중

[논문리뷰] Think, Act, Build: An Agentic Framework with Vision Language Models for Zero-Shot 3D Visual Grounding

본 논문은 3D-VG 작업을 'Think(추론)', 'Act(도구 호출)', 'Build(재구성)' 단계로 세분화한 TAB 프레임워크를 제안합니다 . TAB은 고정된 파이프라인 대신, 전문적인 3D-VG Skill blueprint에 따라 VLM 에이전트가 능동적으로 visual tool을 호출하여 타겟을 추적하고 마스크를 생성합니다.
#Review#3D Visual Grounding#Vision-Language Models#Agentic Framework#RGB-D#Zero-Shot#Geometric Reconstruction
댓글 수 로딩 중

[논문리뷰] Terminal Agents Suffice for Enterprise Automation

저자들은 터미널과 파일시스템을 통해 플랫폼 API와 직접 통신하는 최소한의 코딩 에이전트인 StarShell을 제안합니다 . StarShell은 사전 정의된 도구 레지스트리에 의존하지 않고, 문서나 API 응답을 통해 능동적으로 기능을 발견하고 작업을 구성합니다.
#Review#Enterprise Automation#Agentic Systems#Terminal-based Agents#API Interaction#Model Context Protocol (MCP)#Coding Agents
댓글 수 로딩 중

[논문리뷰] Revision or Re-Solving? Decomposing Second-Pass Gains in Multi-LLM Pipelines

본 논문은 Four-Condition Design을 통해 성능 이득을 Additive하게 분해하는 프레임워크를 제안합니다. 이 방법론은 Generator 기반 성능($x_1$), 표준 Revision($x_2$), 독립 재해결 제어($x_3$), 구조화된 Null 초안 제어($x_4$)를 비교하여 세 가지 효과를 각각 산출합니다.
#Review#Multi-LLM Pipeline#Iterative Refinement#Self-Correction#Task-Time Scaling#Code Generation#MCQ
댓글 수 로딩 중

[논문리뷰] Reasoning Shift: How Context Silently Shortens LLM Reasoning

본 연구는 모델이 동일한 문제를 Baseline, Subtask, Long input, Multi-turn 등 서로 다른 4가지 Context 조건에서 해결하도록 설계하여 추론 성능과 생성된 토큰 수를 측정했습니다. 실험 결과, 모델은 고립된 환경(Baseline) 대비 다른 조건들에서 최대 50%까지 짧은 추론 트레이스를 생성하는 현상을 보였습니다.
#Review#Large Language Models#Chain-of-Thought#Test-time Scaling#Context Management#Reasoning Shift#Self-verification#Overthinking
댓글 수 로딩 중

[논문리뷰] MemRerank: Preference Memory for Personalized Product Reranking

저자들은 사용자 구매 이력을 within-category와 cross-category 선호도로 구분하여 추출하는 Memory Extractor 모델을 설계하였다. 이 추출기는 재순위화 작업의 성능을 극대화하기 위해 GRPO 기반의 강화학습으로 post-training 되며, 이를 통해 별도의 Gold Label 없이도 재순위화 보상을 직접 최적화한다 .
#Review#Recommender Systems#Large Language Models#Product Reranking#Preference Memory#Reinforcement Learning
댓글 수 로딩 중

[논문리뷰] HippoCamp: Benchmarking Contextual Agents on Personal Computers

HippoCamp는 100명 이상의 실제 사용자 데이터를 기반으로 2,000개 이상의 이질적인 파일을 포함하는 3개의 대표적인 개인 컴퓨팅 환경을 구축하였습니다. 각 환경에는 Factual Retention 및 Profiling 태스크를 위한 581개의 QA 쌍이 존재하며, 이를 위해 46.1K개의 정교한 주석(Annotation)이 제공됩니다.
#Review#Multimodal Agents#File-System#Contextual Benchmarking#Personalized Memory#Profiling#Factual Retention
댓글 수 로딩 중

[논문리뷰] AI Generalisation Gap In Comorbid Sleep Disorder Staging

본 연구에서는 SE-ResNet 블록과 Bi-LSTM 레이어를 결합하여 시계열 EEG 데이터의 특징을 추출하고 시간적 의존성을 모델링하는 프레임워크를 제안합니다. 저자들은 새로 구축한 iSLEEPS 데이터셋을 활용하여 기존의 헬스케어 기반 모델들과 성능을 비교하였습니다.
#Review#Sleep Staging#Ischemic Stroke#Explainable AI#Grad-CAM#Deep Learning#Clinical Generalization#iSLEEPS
댓글 수 로딩 중

[논문리뷰] VISion On Request: Enhanced VLLM efficiency with sparse, dynamically selected, vision-language interactions

Georgios Tzimiropoulos이 arXiv에 게시한 'VISion On Request: Enhanced VLLM efficiency with sparse, dynamically selected, vision-language interactions' 논문에 대한 자세한 리뷰입니다.
#Review#LVLM Efficiency#Sparse Interaction#Cross-Attention#Self-Attention#Adaptive Inference#Visual Feature Refinement#Computational Cost Reduction
댓글 수 로딩 중

[논문리뷰] Attend Before Attention: Efficient and Scalable Video Understanding via Autoregressive Gazing

David Eigen이 arXiv에 게시한 'Attend Before Attention: Efficient and Scalable Video Understanding via Autoregressive Gazing' 논문에 대한 자세한 리뷰입니다.
#Review#Video Understanding#Multi-modal Large Language Models (MLLMs)#Vision Transformers (ViTs)#Autoregressive Gazing#Token Reduction#Multi-scale Patches#High-Resolution Video#Long-Form Video
댓글 수 로딩 중

[논문리뷰] VideoDetective: Clue Hunting via both Extrinsic Query and Intrinsic Relevance for Long Video Understanding

arXiv에 게시된 'VideoDetective: Clue Hunting via both Extrinsic Query and Intrinsic Relevance for Long Video Understanding' 논문에 대한 자세한 리뷰입니다.
#Review#Long Video Understanding#Multimodal Large Language Models#Video Question Answering#Graph Neural Networks#Active Inference#Belief Propagation#Spatio-Temporal Graph
댓글 수 로딩 중

[논문리뷰] RoboAlign: Learning Test-Time Reasoning for Language-Action Alignment in Vision-Language-Action Models

arXiv에 게시된 'RoboAlign: Learning Test-Time Reasoning for Language-Action Alignment in Vision-Language-Action Models' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language-Action Models (VLAs)#Multimodal-Large-Language Models (MLLMs)#Reinforcement Learning (RL)#Supervised Fine-tuning (SFT)#Embodied Reasoning#Low-level Actions#FAST tokenization#Robotics
댓글 수 로딩 중

[논문리뷰] LongCat-Flash-Prover: Advancing Native Formal Reasoning via Agentic Tool-Integrated Reinforcement Learning

arXiv에 게시된 'LongCat-Flash-Prover: Advancing Native Formal Reasoning via Agentic Tool-Integrated Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Mixture-of-Experts#Native Formal Reasoning#Tool-Integrated Reinforcement Learning#Lean4#Auto-formalization#Theorem Proving#Hierarchical Importance Sampling Policy Optimization
댓글 수 로딩 중

[논문리뷰] Astrolabe: Steering Forward-Process Reinforcement Learning for Distilled Autoregressive Video Models

Jie Huang이 arXiv에 게시한 'Astrolabe: Steering Forward-Process Reinforcement Learning for Distilled Autoregressive Video Models' 논문에 대한 자세한 리뷰입니다.
#Review#Video Generation#Distilled Autoregressive Models#Reinforcement Learning (RL)#Human Preferences#Streaming Generation#Forward-Process RL#Reward Hacking#Temporal Consistency
댓글 수 로딩 중

[논문리뷰] Think While Watching: Online Streaming Segment-Level Memory for Multi-Turn Video Reasoning in Multimodal Large Language Models

Kang Liu이 arXiv에 게시한 'Think While Watching: Online Streaming Segment-Level Memory for Multi-Turn Video Reasoning in Multimodal Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Streaming Video Reasoning#Multi-Turn Interaction#Segment-Level Memory#Causal Mask#Positional Encoding#Dual KV Cache#Multimodal Large Language Models
댓글 수 로딩 중

[논문리뷰] HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarios

arXiv에 게시된 'HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarios' 논문에 대한 자세한 리뷰입니다.
#Review#Embodied Agents#Unsafe Action Detection#Vision-Language Models (VLMs)#Household Scenarios#HomeSafe-Bench#HD-Guard#Real-time Safety Monitoring
댓글 수 로딩 중

[논문리뷰] Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation

Yichen Zhang이 arXiv에 게시한 'Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Unified multimodal model#Visual generation and comprehension#Unified vision encoder#Cascaded flow matching#Token compression
댓글 수 로딩 중

[논문리뷰] MiniAppBench: Evaluating the Shift from Text to Interactive HTML Responses in LLM-Powered Assistants

Yuante Li이 arXiv에 게시한 'MiniAppBench: Evaluating the Shift from Text to Interactive HTML Responses in LLM-Powered Assistants' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models (LLMs)#Code Generation#HTML#Interactive Applications#Benchmark#MINIAPPBENCH#Agentic Evaluation#MINIAPPEVAL#Real-World Principles#Human-AI Interaction
댓글 수 로딩 중

[논문리뷰] Skip to the Good Part: Representation Structure & Inference-Time Layer Skipping in Diffusion vs. Autoregressive LLMs

arXiv에 게시된 'Skip to the Good Part: Representation Structure & Inference-Time Layer Skipping in Diffusion vs. Autoregressive LLMs' 논문에 대한 자세한 리뷰입니다.
#Review#Diffusion LLMs#Autoregressive LLMs#Representational Analysis#Layer Skipping#Inference Efficiency#Initialization Bias#FLOPs Reduction#Recency Bias
댓글 수 로딩 중

[논문리뷰] STMI: Segmentation-Guided Token Modulation with Cross-Modal Hypergraph Interaction for Multi-Modal Object Re-Identification

arXiv에 게시된 'STMI: Segmentation-Guided Token Modulation with Cross-Modal Hypergraph Interaction for Multi-Modal Object Re-Identification' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-modal Re-Identification#Segmentation-Guided Feature Modulation#Token Modulation#Cross-Modal Interaction#Hypergraph Neural Networks#Object ReID#Transformer#SAM
댓글 수 로딩 중

[논문리뷰] HiFi-Inpaint: Towards High-Fidelity Reference-Based Inpainting for Generating Detail-Preserving Human-Product Images

arXiv에 게시된 'HiFi-Inpaint: Towards High-Fidelity Reference-Based Inpainting for Generating Detail-Preserving Human-Product Images' 논문에 대한 자세한 리뷰입니다.
#Review#Reference-Based Inpainting#High-Fidelity Image Generation#Human-Product Images#Diffusion Models#Detail Preservation#Attention Mechanisms#Loss Functions#Dataset Construction
댓글 수 로딩 중

[논문리뷰] BeamPERL: Parameter-Efficient RL with Verifiable Rewards Specializes Compact LLMs for Structured Beam Mechanics Reasoning

arXiv에 게시된 'BeamPERL: Parameter-Efficient RL with Verifiable Rewards Specializes Compact LLMs for Structured Beam Mechanics Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Parameter-Efficient Fine-Tuning (PEFT)#Large Language Models (LLM)#Beam Mechanics#Verifiable Rewards#Engineering Reasoning#Structural Engineering#Group Relative Policy Optimization (GRPO)
댓글 수 로딩 중

[논문리뷰] Reasoning Core: A Scalable Procedural Data Generation Suite for Symbolic Pre-training and Post-Training

arXiv에 게시된 'Reasoning Core: A Scalable Procedural Data Generation Suite for Symbolic Pre-training and Post-Training' 논문에 대한 자세한 리뷰입니다.
#Review#Procedural Data Generation#Symbolic Reasoning#Language Model Pre-training#Reinforcement Learning with Verifiable Rewards#Formal Logic#PDDL Planning#Context-Free Grammars
댓글 수 로딩 중

[논문리뷰] AI Gamestore: Scalable, Open-Ended Evaluation of Machine General Intelligence with Human Games

arXiv에 게시된 'AI Gamestore: Scalable, Open-Ended Evaluation of Machine General Intelligence with Human Games' 논문에 대한 자세한 리뷰입니다.
#Review#Artificial General Intelligence (AGI)#Evaluation Benchmark#General Game Playing#Large Language Models (LLMs)#Human-in-the-loop#Cognitive Capabilities#Vision-Language Models (VLMs)#Game Generation
댓글 수 로딩 중

[논문리뷰] NoLan: Mitigating Object Hallucinations in Large Vision-Language Models via Dynamic Suppression of Language Priors

Xinchao Wang이 arXiv에 게시한 'NoLan: Mitigating Object Hallucinations in Large Vision-Language Models via Dynamic Suppression of Language Priors' 논문에 대한 자세한 리뷰입니다.
#Review#Large Vision-Language Models (LVLMs)#Object Hallucinations#Language Priors#Contrastive Decoding#Dynamic Suppression#Training-Free#Multimodal AI
댓글 수 로딩 중

[논문리뷰] Model Context Protocol (MCP) Tool Descriptions Are Smelly! Towards Improving AI Agent Efficiency with Augmented MCP Tool Descriptions

Ahmed E. Hassan이 arXiv에 게시한 'Model Context Protocol (MCP) Tool Descriptions Are Smelly! Towards Improving AI Agent Efficiency with Augmented MCP Tool Descriptions' 논문에 대한 자세한 리뷰입니다.
#Review#Model Context Protocol#AI Agents#Tool Descriptions#Software Smells#Prompt Engineering#Foundation Models#Performance Evaluation#Ablation Study
댓글 수 로딩 중

[논문리뷰] FlowPrefill: Decoupling Preemption from Prefill Scheduling Granularity to Mitigate Head-of-Line Blocking in LLM Serving

Jidong Zhai이 arXiv에 게시한 'FlowPrefill: Decoupling Preemption from Prefill Scheduling Granularity to Mitigate Head-of-Line Blocking in LLM Serving' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Serving#Head-of-Line Blocking#Preemption#Prefill Scheduling#Time-to-First-Token (TTFT)#SLO-aware Scheduling#Operator-Level Preemption#Event-Driven Scheduling
댓글 수 로딩 중

[논문리뷰] Conv-FinRe: A Conversational and Longitudinal Benchmark for Utility-Grounded Financial Recommendation

arXiv에 게시된 'Conv-FinRe: A Conversational and Longitudinal Benchmark for Utility-Grounded Financial Recommendation' 논문에 대한 자세한 리뷰입니다.
#Review#Financial Recommendation#Conversational AI#Large Language Models#Utility-Grounded Evaluation#Behavioral Finance#Stock Recommendation#Longitudinal Benchmark#Inverse Optimization
댓글 수 로딩 중

[논문리뷰] ManCAR: Manifold-Constrained Latent Reasoning with Adaptive Test-Time Computation for Sequential Recommendation

arXiv에 게시된 'ManCAR: Manifold-Constrained Latent Reasoning with Adaptive Test-Time Computation for Sequential Recommendation' 논문에 대한 자세한 리뷰입니다.
#Review#Sequential Recommendation#Latent Reasoning#Manifold Constraint#Adaptive Computation#Graph Neural Networks#Variational Inference#Teacher Scheduling#Drift Prevention
댓글 수 로딩 중

[논문리뷰] Generated Reality: Human-centric World Simulation using Interactive Video Generation with Hand and Camera Control

Shengqu Cai이 arXiv에 게시한 'Generated Reality: Human-centric World Simulation using Interactive Video Generation with Hand and Camera Control' 논문에 대한 자세한 리뷰입니다.
#Review#Video Generation#Extended Reality (XR)#Diffusion Models#Human-Computer Interaction#Hand Pose Estimation#Camera Control#World Simulation#Interactive AI
댓글 수 로딩 중

[논문리뷰] EgoPush: Learning End-to-End Egocentric Multi-Object Rearrangement for Mobile Robots

Sihang Li이 arXiv에 게시한 'EgoPush: Learning End-to-End Egocentric Multi-Object Rearrangement for Mobile Robots' 논문에 대한 자세한 리뷰입니다.
#Review#Egocentric Perception#Multi-Object Rearrangement#Mobile Robotics#Reinforcement Learning#Teacher-Student Distillation#Non-Prehensile Manipulation#Sim-to-Real Transfer#Object-Centric Representation
댓글 수 로딩 중

[논문리뷰] InnoEval: On Research Idea Evaluation as a Knowledge-Grounded, Multi-Perspective Reasoning Problem

arXiv에 게시된 'InnoEval: On Research Idea Evaluation as a Knowledge-Grounded, Multi-Perspective Reasoning Problem' 논문에 대한 자세한 리뷰입니다.
#Review#Research Idea Evaluation#Large Language Models (LLMs)#Knowledge Grounding#Multi-Perspective Reasoning#Agent-based Systems#Scientific Discovery#Peer Review Simulation#Automated Evaluation
댓글 수 로딩 중

[논문리뷰] Acoustivision Pro: An Open-Source Interactive Platform for Room Impulse Response Analysis and Acoustic Characterization

Mandip Goswami이 arXiv에 게시한 'Acoustivision Pro: An Open-Source Interactive Platform for Room Impulse Response Analysis and Acoustic Characterization' 논문에 대한 자세한 리뷰입니다.
#Review#Room Acoustics#Room Impulse Response#Acoustic Analysis#Open-Source Platform#Web Application#Data Visualization#Acoustic Metrics#Standards Compliance
댓글 수 로딩 중

[논문리뷰] Recurrent-Depth VLA: Implicit Test-Time Compute Scaling of Vision-Language-Action Models via Latent Iterative Reasoning

arXiv에 게시된 'Recurrent-Depth VLA: Implicit Test-Time Compute Scaling of Vision-Language-Action Models via Latent Iterative Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language-Action Models#Latent Iterative Reasoning#Adaptive Compute#Recurrent Neural Networks#Robotics#Transformer#Test-Time Scaling#Continuous Action Space
댓글 수 로딩 중

[논문리뷰] Judging What We Cannot Solve: A Consequence-Based Approach for Oracle-Free Evaluation of Research-Level Math

Amit Agarwal이 arXiv에 게시한 'Judging What We Cannot Solve: A Consequence-Based Approach for Oracle-Free Evaluation of Research-Level Math' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Evaluation#Mathematical Reasoning#Oracle-Free Validation#Consequence-Based Utility#Solution Quality#In-Context Learning#Research-Level Math
댓글 수 로딩 중

[논문리뷰] Back to Basics: Revisiting Exploration in Reinforcement Learning for LLM Reasoning via Generative Probabilities

Ivan Oseledets이 arXiv에 게시한 'Back to Basics: Revisiting Exploration in Reinforcement Learning for LLM Reasoning via Generative Probabilities' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#LLM Reasoning#Exploration-Exploitation#Group Relative Policy Optimization#Entropy Collapse#Generative Models#Confidence-Aware Rewards
댓글 수 로딩 중

[논문리뷰] Length-Unbiased Sequence Policy Optimization: Revealing and Controlling Response Length Variation in RLVR

Zhixiong Zeng이 arXiv에 게시한 'Length-Unbiased Sequence Policy Optimization: Revealing and Controlling Response Length Variation in RLVR' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning with Verifiable Rewards#LLMs#Policy Optimization#Response Length Bias#Sequence-level Clipping#Length-Unbiased Optimization#Multimodal Reasoning
댓글 수 로딩 중

[논문리뷰] WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning

arXiv에 게시된 'WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-Agent Reinforcement Learning#Width Scaling#Large Language Models#Information Seeking#Task Decomposition#Parallel Execution#Lead-Agent-Subagent Framework#Orchestration
댓글 수 로딩 중

[논문리뷰] EgoActor: Grounding Task Planning into Spatial-aware Egocentric Actions for Humanoid Robots via Visual-Language Models

Ziyi Bai이 arXiv에 게시한 'EgoActor: Grounding Task Planning into Spatial-aware Egocentric Actions for Humanoid Robots via Visual-Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Humanoid Robots#Vision-Language Models#Task Planning#Egocentric Control#Mobile Manipulation#Active Perception#Human-Robot Interaction#Real-World Deployment
댓글 수 로딩 중

[논문리뷰] BatCoder: Self-Supervised Bidirectional Code-Documentation Learning via Back-Translation

Xiaohua Wang이 arXiv에 게시한 'BatCoder: Self-Supervised Bidirectional Code-Documentation Learning via Back-Translation' 논문에 대한 자세한 리뷰입니다.
#Review#Self-Supervised Learning#Code Generation#Documentation Generation#Back-Translation#Reinforcement Learning#Large Language Models (LLMs)#Code-Documentation Alignment#Low-Resource Languages
댓글 수 로딩 중

[논문리뷰] AdaptMMBench: Benchmarking Adaptive Multimodal Reasoning for Mode Selection and Reasoning Process

Shilin Yan이 arXiv에 게시한 'AdaptMMBench: Benchmarking Adaptive Multimodal Reasoning for Mode Selection and Reasoning Process' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Reasoning#Adaptive Learning#Vision-Language Models (VLMs)#Benchmarking#Mode Selection#Tool Learning#Reasoning Process Evaluation#Matthews Correlation Coefficient (MCC)
댓글 수 로딩 중

[논문리뷰] How Well Do Models Follow Visual Instructions? VIBE: A Systematic Benchmark for Visual Instruction-Driven Image Editing

Haochen Tian이 arXiv에 게시한 'How Well Do Models Follow Visual Instructions? VIBE: A Systematic Benchmark for Visual Instruction-Driven Image Editing' 논문에 대한 자세한 리뷰입니다.
#Review#Visual Instruction#Image Editing#Multimodal Benchmark#LMM-as-a-judge#Deictic Grounding#Morphological Manipulation#Causal Reasoning#Generative Models
댓글 수 로딩 중

[논문리뷰] Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation

arXiv에 게시된 'Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Autoregressive Video Generation#Diffusion Models#Model Distillation#Real-Time AI#Causal Attention#ODE Distillation#Frame-level Injectivity#Teacher Forcing
댓글 수 로딩 중

[논문리뷰] FourierSampler: Unlocking Non-Autoregressive Potential in Diffusion Language Models via Frequency-Guided Generation

arXiv에 게시된 'FourierSampler: Unlocking Non-Autoregressive Potential in Diffusion Language Models via Frequency-Guided Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Diffusion Language Models#Non-Autoregressive Generation#Frequency Domain Analysis#Decoding Strategy#Structure-to-Detail#Fourier Transform#Text Generation
댓글 수 로딩 중

[논문리뷰] Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening

Haitham Bou Ammar이 arXiv에 게시한 'Scalable Power Sampling: Unlocking Efficient, Training-Free Reasoning for LLMs via Distribution Sharpening' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Reasoning#Distribution Sharpening#Power Sampling#Training-Free#Monte Carlo Estimation#Jackknife Correction#Autoregressive Generation#Inference Efficiency
댓글 수 로딩 중

[논문리뷰] MAD: Modality-Adaptive Decoding for Mitigating Cross-Modal Hallucinations in Multimodal Large Language Models

Yong Man Ro이 arXiv에 게시한 'MAD: Modality-Adaptive Decoding for Mitigating Cross-Modal Hallucinations in Multimodal Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal LLM#Cross-modal Hallucination#Contrastive Decoding#Modality-Adaptive Decoding#Self-Assessment#Audio-Visual Language Model#Training-Free
댓글 수 로딩 중

[논문리뷰] SketchDynamics: Exploring Free-Form Sketches for Dynamic Intent Expression in Animation Generation

Hongbo Fu이 arXiv에 게시한 'SketchDynamics: Exploring Free-Form Sketches for Dynamic Intent Expression in Animation Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Animation Generation#Free-Form Sketching#Human-AI Interaction#Vision-Language Models (VLMs)#Dynamic Intent Expression#Motion Graphics#Iterative Refinement#Storyboard
댓글 수 로딩 중

[논문리뷰] HalluCitation Matters: Revealing the Impact of Hallucinated References with 300 Hallucinated Papers in ACL Conferences

Taro Watanabe이 arXiv에 게시한 'HalluCitation Matters: Revealing the Impact of Hallucinated References with 300 Hallucinated Papers in ACL Conferences' 논문에 대한 자세한 리뷰입니다.
#Review#Hallucinated Citations#NLP Conferences#Citation Detection#Academic Integrity#Peer Review#Large Language Models (LLMs)#Bibliometrics
댓글 수 로딩 중

[논문리뷰] Less Is More -- Until It Breaks: Security Pitfalls of Vision Token Compression in Large Vision-Language Models

Guanhong Tao이 arXiv에 게시한 'Less Is More -- Until It Breaks: Security Pitfalls of Vision Token Compression in Large Vision-Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#LVLM Security#Token Compression#Adversarial Attack#Robustness Degradation#Compression-Aware Attack#Efficiency-Security Trade-off#Black-box Attack
댓글 수 로딩 중

[논문리뷰] TwinBrainVLA: Unleashing the Potential of Generalist VLMs for Embodied Tasks via Asymmetric Mixture-of-Transformers

arXiv에 게시된 'TwinBrainVLA: Unleashing the Potential of Generalist VLMs for Embodied Tasks via Asymmetric Mixture-of-Transformers' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language-Action (VLA)#Embodied AI#Robotics#Catastrophic Forgetting#Asymmetric Mixture-of-Transformers (AsyMoT)#Generalist VLM#Specialist VLM#Flow-Matching
댓글 수 로딩 중

[논문리뷰] Inference-Time Scaling of Verification: Self-Evolving Deep Research Agents via Test-Time Rubric-Guided Verification

arXiv에 게시된 'Inference-Time Scaling of Verification: Self-Evolving Deep Research Agents via Test-Time Rubric-Guided Verification' 논문에 대한 자세한 리뷰입니다.
#Review#Deep Research Agents#Inference-Time Verification#Self-Evolving LLM Agents#Rubric-Guided Feedback#Failure Taxonomy#Test-Time Scaling#Supervised Fine-tuning
댓글 수 로딩 중

[논문리뷰] Facilitating Proactive and Reactive Guidance for Decision Making on the Web: A Design Probe with WebSeek

Arpit Narechania이 arXiv에 게시한 'Facilitating Proactive and Reactive Guidance for Decision Making on the Web: A Design Probe with WebSeek' 논문에 대한 자세한 리뷰입니다.
#Review#Mixed-Initiative AI#Human-AI Collaboration#Web Data Analysis#Proactive Guidance#Large Language Models (LLMs)#Browser Extension#Data-Centric Design
댓글 수 로딩 중

[논문리뷰] A Hybrid Protocol for Large-Scale Semantic Dataset Generation in Low-Resource Languages: The Turkish Semantic Relations Corpus

Özay Ezerceli이 arXiv에 게시한 'A Hybrid Protocol for Large-Scale Semantic Dataset Generation in Low-Resource Languages: The Turkish Semantic Relations Corpus' 논문에 대한 자세한 리뷰입니다.
#Review#Low-Resource NLP#Semantic Relations#Dataset Generation#Turkish Language#LLM#FastText Embeddings#Agglomerative Clustering#Synonyms#Antonyms#Co-hyponyms
댓글 수 로딩 중

[논문리뷰] A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5

Yutao Wu이 arXiv에 게시한 'A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Doubao 1.8, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5' 논문에 대한 자세한 리뷰입니다.
#Review#AI Safety#Large Language Models#Multimodal LLMs#Benchmark Evaluation#Adversarial Robustness#Multilingual Evaluation#Regulatory Compliance#Image Generation Safety
댓글 수 로딩 중

[논문리뷰] SkinFlow: Efficient Information Transmission for Open Dermatological Diagnosis via Dynamic Visual Encoding and Staged RL

arXiv에 게시된 'SkinFlow: Efficient Information Transmission for Open Dermatological Diagnosis via Dynamic Visual Encoding and Staged RL' 논문에 대한 자세한 리뷰입니다.
#Review#Dermatological Diagnosis#Multimodal LLM#Reinforcement Learning#Dynamic Visual Encoding#Information Transmission#Clinically Grounded Evaluation
댓글 수 로딩 중

[논문리뷰] OpenVoxel: Training-Free Grouping and Captioning Voxels for Open-Vocabulary 3D Scene Understanding

arXiv에 게시된 'OpenVoxel: Training-Free Grouping and Captioning Voxels for Open-Vocabulary 3D Scene Understanding' 논문에 대한 자세한 리뷰입니다.
#Review#3D Scene Understanding#Open-Vocabulary Segmentation#Referring Expression Segmentation#Training-Free#Voxel Grouping#Vision-Language Models#Multi-modal Large Language Models#Sparse Voxel Rasterization
댓글 수 로딩 중

[논문리뷰] Are LLMs Vulnerable to Preference-Undermining Attacks (PUA)? A Factorial Analysis Methodology for Diagnosing the Trade-off between Preference Alignment and Real-World Validity

Chi Zhang이 arXiv에 게시한 'Are LLMs Vulnerable to Preference-Undermining Attacks (PUA)? A Factorial Analysis Methodology for Diagnosing the Trade-off between Preference Alignment and Real-World Validity' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models#Preference Alignment#Preference-Undermining Attacks#Factorial Analysis#Sycophancy#Prompt Engineering#Truth-Deference Trade-off
댓글 수 로딩 중

[논문리뷰] Controllable Memory Usage: Balancing Anchoring and Innovation in Long-Term Human-Agent Interaction

Zhengkang Guo이 arXiv에 게시한 'Controllable Memory Usage: Balancing Anchoring and Innovation in Long-Term Human-Agent Interaction' 논문에 대한 자세한 리뷰입니다.
#Review#Long-Term Human-Agent Interaction#Controllable Memory#Memory Anchoring#Large Language Models (LLMs)#Personalization#Reinforcement Learning (RL)#Supervised Fine-Tuning (SFT)#Memory Dependence
댓글 수 로딩 중

[논문리뷰] Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking

arXiv에 게시된 'Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Retrieval#Multimodal Ranking#Foundation Models#Embedding Models#Reranking Models#Contrastive Learning#Knowledge Distillation#Matryoshka Representation Learning#Quantization-Aware Training
댓글 수 로딩 중

[논문리뷰] RL-AWB: Deep Reinforcement Learning for Auto White Balance Correction in Low-Light Night-time Scenes

Yu-Lun Liu이 arXiv에 게시한 'RL-AWB: Deep Reinforcement Learning for Auto White Balance Correction in Low-Light Night-time Scenes' 논문에 대한 자세한 리뷰입니다.
#Review#Auto White Balance (AWB)#Deep Reinforcement Learning (DRL)#Low-Light Imaging#Night-time Scenes#Color Constancy#Cross-Sensor Generalization#Statistical Methods#Curriculum Learning
댓글 수 로딩 중

[논문리뷰] EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering for Enhanced Alignment and Reasoning

Guanchen Wu이 arXiv에 게시한 'EpiQAL: Benchmarking Large Language Models in Epidemiological Question Answering for Enhanced Alignment and Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Epidemiological Question Answering#Large Language Models#Benchmark#Multi-step Inference#Evidence Grounding#LLM Evaluation#Public Health AI#Chain-of-Thought
댓글 수 로딩 중

[논문리뷰] X-MuTeST: A Multilingual Benchmark for Explainable Hate Speech Detection and A Novel LLM-consulted Explanation Framework

Shwetank Shekhar Singh이 arXiv에 게시한 'X-MuTeST: A Multilingual Benchmark for Explainable Hate Speech Detection and A Novel LLM-consulted Explanation Framework' 논문에 대한 자세한 리뷰입니다.
#Review#Hate Speech Detection#Explainable AI (XAI)#Multilingual NLP#Large Language Models (LLMs)#Attention Mechanism#N-gram Explanations#Human Rationales#Benchmark Dataset
댓글 수 로딩 중

[논문리뷰] CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solving

Tao Feng이 arXiv에 게시한 'CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solving' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal LLMs#Visual Reasoning#Mathematical Problem Solving#Knowledge Internalization#Reinforcement Learning#Cognitive-Inspired AI#Perception-Reasoning Alignment
댓글 수 로딩 중

[논문리뷰] Talk2Move: Reinforcement Learning for Text-Instructed Object-Level Geometric Transformation in Scenes

Shuo Yang이 arXiv에 게시한 'Talk2Move: Reinforcement Learning for Text-Instructed Object-Level Geometric Transformation in Scenes' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Text-Guided Image Editing#Object-Level Transformation#Geometric Transformation#Diffusion Models#GRPO#Scene Editing#Spatially Grounded Rewards
댓글 수 로딩 중

[논문리뷰] Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem

Wei Gao이 arXiv에 게시한 'Let It Flow: Agentic Crafting on Rock and Roll, Building the ROME Model within an Open Agentic Learning Ecosystem' 논문에 대한 자세한 리뷰입니다.
#Review#Agentic Learning Ecosystem#Large Language Models#Reinforcement Learning#Agentic Crafting#Tool Use#ROME Model#Policy Optimization#Sandbox Environment
댓글 수 로딩 중

[논문리뷰] Geometry-Aware Optimization for Respiratory Sound Classification: Enhancing Sensitivity with SAM-Optimized Audio Spectrogram Transformers

Mahşuk Taylan이 arXiv에 게시한 'Geometry-Aware Optimization for Respiratory Sound Classification: Enhancing Sensitivity with SAM-Optimized Audio Spectrogram Transformers' 논문에 대한 자세한 리뷰입니다.
#Review#Respiratory Sound Classification#Audio Spectrogram Transformer#Sharpness-Aware Minimization#Loss Landscape#Imbalanced Learning#Transfer Learning#ICBHI 2017
댓글 수 로딩 중

[논문리뷰] Fantastic Reasoning Behaviors and Where to Find Them: Unsupervised Discovery of the Reasoning Process

arXiv에 게시된 'Fantastic Reasoning Behaviors and Where to Find Them: Unsupervised Discovery of the Reasoning Process' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Reasoning#Mechanistic Interpretability#Sparse Autoencoders (SAEs)#Activation Steering#Unsupervised Learning#Reasoning Behaviors#Chain-of-Thought#Feature Disentanglement
댓글 수 로딩 중

[논문리뷰] Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone

arXiv에 게시된 'Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone' 논문에 대한 자세한 리뷰입니다.
#Review#Diffusion Models#Vision-Language Models#Vision-Language-Action Models#Robotics#Multimodal AI#Action Planning#Long-Horizon Planning#Bidirectional Attention
댓글 수 로딩 중

[논문리뷰] UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture

Kaiwen Zhu이 arXiv에 게시한 'UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture' 논문에 대한 자세한 리뷰입니다.
#Review#Perceptual Understanding#Image Aesthetics#Image Quality#Image Structure#Image Texture#MLLM Benchmark#Visual Question Answering#Reward Model
댓글 수 로딩 중

[논문리뷰] Beyond Memorization: A Multi-Modal Ordinal Regression Benchmark to Expose Popularity Bias in Vision-Language Models

Yu-Lun Liu이 arXiv에 게시한 'Beyond Memorization: A Multi-Modal Ordinal Regression Benchmark to Expose Popularity Bias in Vision-Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language Models (VLMs)#Popularity Bias#Ordinal Regression#Building Age Estimation#Multi-modal Learning#Benchmark Dataset#Explainable AI
댓글 수 로딩 중

[논문리뷰] Simulstream: Open-Source Toolkit for Evaluation and Demonstration of Streaming Speech-to-Text Translation Systems

Luisa Bentivogli이 arXiv에 게시한 'Simulstream: Open-Source Toolkit for Evaluation and Demonstration of Streaming Speech-to-Text Translation Systems' 논문에 대한 자세한 리뷰입니다.
#Review#Streaming Speech-to-Text Translation#StreamST#Evaluation Toolkit#Open-Source Framework#Re-translation#Incremental Decoding#Latency Metrics#Quality Metrics#Real-time Demonstration
댓글 수 로딩 중

[논문리뷰] Multi-LLM Thematic Analysis with Dual Reliability Metrics: Combining Cohen's Kappa and Semantic Similarity for Qualitative Research Validation

arXiv에 게시된 'Multi-LLM Thematic Analysis with Dual Reliability Metrics: Combining Cohen's Kappa and Semantic Similarity for Qualitative Research Validation' 논문에 대한 자세한 리뷰입니다.
#Review#Thematic Analysis#Large Language Models#Qualitative Research#Cohen's Kappa#Semantic Similarity#Reliability Metrics#Ensemble Validation#Prompt Engineering
댓글 수 로딩 중

[논문리뷰] Reasoning Palette: Modulating Reasoning via Latent Contextualization for Controllable Exploration for (V)LMs

arXiv에 게시된 'Reasoning Palette: Modulating Reasoning via Latent Contextualization for Controllable Exploration for (V)LMs' 논문에 대한 자세한 리뷰입니다.
#Review#Latent Variable Models#Variational Autoencoder (VAE)#Reinforcement Learning (RL)#Exploration#Large Language Models (LLMs)#Vision-Language Models (VLMs)#Controllable Generation#Reasoning Strategies
댓글 수 로딩 중

[논문리뷰] MobileWorld: Benchmarking Autonomous Mobile Agents in Agent-User Interactive, and MCP-Augmented Environments

arXiv에 게시된 'MobileWorld: Benchmarking Autonomous Mobile Agents in Agent-User Interactive, and MCP-Augmented Environments' 논문에 대한 자세한 리뷰입니다.
#Review#Mobile Agents#GUI Benchmarking#Agent-User Interaction#Tool-Augmented Agents#Model Context Protocol (MCP)#Long-Horizon Tasks#Reproducible Evaluation#Android Environment
댓글 수 로딩 중

[논문리뷰] Can LLMs Estimate Student Struggles? Human-AI Difficulty Alignment with Proficiency Simulation for Item Difficulty Prediction

Hong Jiao이 arXiv에 게시한 'Can LLMs Estimate Student Struggles? Human-AI Difficulty Alignment with Proficiency Simulation for Item Difficulty Prediction' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models#Item Difficulty Prediction#Human-AI Alignment#Proficiency Simulation#Metacognition#Curse of Knowledge#Educational Assessment#Zero-shot Learning
댓글 수 로딩 중

[논문리뷰] SWE-Bench++: A Framework for the Scalable Generation of Software Engineering Benchmarks from Open-Source Repositories

arXiv에 게시된 'SWE-Bench++: A Framework for the Scalable Generation of Software Engineering Benchmarks from Open-Source Repositories' 논문에 대한 자세한 리뷰입니다.
#Review#Software Engineering Benchmarks#Large Language Models (LLMs)#Code Generation#Automated Benchmark Generation#Multilingual#GitHub Pull Requests#Test Oracle#Fine-tuning
댓글 수 로딩 중

[논문리뷰] Both Semantics and Reconstruction Matter: Making Representation Encoders Ready for Text-to-Image Generation and Editing

arXiv에 게시된 'Both Semantics and Reconstruction Matter: Making Representation Encoders Ready for Text-to-Image Generation and Editing' 논문에 대한 자세한 리뷰입니다.
#Review#Text-to-Image Generation#Image Editing#Representation Encoders#Latent Diffusion Models#Variational Autoencoder (VAE)#Semantic Reconstruction#Off-manifold Latents#DINOv2
댓글 수 로딩 중

[논문리뷰] ShowTable: Unlocking Creative Table Visualization with Collaborative Reflection and Refinement

Zhaohe Liao이 arXiv에 게시한 'ShowTable: Unlocking Creative Table Visualization with Collaborative Reflection and Refinement' 논문에 대한 자세한 리뷰입니다.
#Review#Table Visualization#Infographic Generation#Multi-modal Large Language Models (MLLMs)#Diffusion Models#Self-Correction#Reinforcement Learning#Graphic Design#Data-to-Visual Mapping
댓글 수 로딩 중

[논문리뷰] H2R-Grounder: A Paired-Data-Free Paradigm for Translating Human Interaction Videos into Physically Grounded Robot Videos

Mike Zheng Shou이 arXiv에 게시한 'H2R-Grounder: A Paired-Data-Free Paradigm for Translating Human Interaction Videos into Physically Grounded Robot Videos' 논문에 대한 자세한 리뷰입니다.
#Review#Video-to-Video Translation#Robot Learning#Human-Robot Transfer#Diffusion Models#Unpaired Data Learning#Pose-Guided Generation#Embodiment Gap Bridging
댓글 수 로딩 중

[논문리뷰] InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models

arXiv에 게시된 'InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language Models#Linear Attention#Sliding Window Attention#Gated DeltaNet#Long-Context Understanding#Efficiency#Hybrid Architecture#Multimodal Learning
댓글 수 로딩 중

[논문리뷰] IF-Bench: Benchmarking and Enhancing MLLMs for Infrared Images with Generative Visual Prompting

arXiv에 게시된 'IF-Bench: Benchmarking and Enhancing MLLMs for Infrared Images with Generative Visual Prompting' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Large Language Models (MLLMs)#Infrared Image Understanding#Benchmark Dataset#Visual Question Answering (VQA)#Generative Visual Prompting (GenViP)#Domain Adaptation#Image-to-Image Translation
댓글 수 로딩 중

[논문리뷰] SUCCESS-GS: Survey of Compactness and Compression for Efficient Static and Dynamic Gaussian Splatting

Sung-Ho Bae이 arXiv에 게시한 'SUCCESS-GS: Survey of Compactness and Compression for Efficient Static and Dynamic Gaussian Splatting' 논문에 대한 자세한 리뷰입니다.
#Review#3D Gaussian Splatting (3DGS)#Gaussian Compression#Model Efficiency#Novel View Synthesis#Dynamic Scenes#Parameter Compression#Restructuring Compression#Real-time Rendering
댓글 수 로딩 중

[논문리뷰] VG-Refiner: Towards Tool-Refined Referring Grounded Reasoning via Agentic Reinforcement Learning

Yansong Tang이 arXiv에 게시한 'VG-Refiner: Towards Tool-Refined Referring Grounded Reasoning via Agentic Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Tool-integrated Visual Reasoning#Referring Grounded Reasoning#Agentic Reinforcement Learning#Self-Correction#Large Vision-Language Models#Chain-of-Thought#Tool Refinement
댓글 수 로딩 중

[논문리뷰] Decouple to Generalize: Context-First Self-Evolving Learning for Data-Scarce Vision-Language Reasoning

arXiv에 게시된 'Decouple to Generalize: Context-First Self-Evolving Learning for Data-Scarce Vision-Language Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language Models#Reinforcement Learning#Self-Evolving Learning#Data-Scarce Domains#Context-First Learning#Reward Hacking Mitigation#Multimodal Reasoning#Curriculum Learning
댓글 수 로딩 중

[논문리뷰] Beyond Token-level Supervision: Unlocking the Potential of Decoding-based Regression via Reinforcement Learning

Jiacheng Chen이 arXiv에 게시한 'Beyond Token-level Supervision: Unlocking the Potential of Decoding-based Regression via Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Decoding-based Regression#Reinforcement Learning#Numerical Prediction#Large Language Models#Policy Gradient#Tokenization#Sequence Generation
댓글 수 로딩 중

[논문리뷰] From Imitation to Discrimination: Toward A Generalized Curriculum Advantage Mechanism Enhancing Cross-Domain Reasoning Tasks

Yang Li이 arXiv에 게시한 'From Imitation to Discrimination: Toward A Generalized Curriculum Advantage Mechanism Enhancing Cross-Domain Reasoning Tasks' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Large Language Models#Curriculum Learning#Advantage Function#Reasoning Tasks#Multimodal AI#Policy Optimization#Generalization
댓글 수 로딩 중

[논문리뷰] SignRoundV2: Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs

arXiv에 게시된 'SignRoundV2: Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs' 논문에 대한 자세한 리뷰입니다.
#Review#Post-Training Quantization (PTQ)#Large Language Models (LLMs)#Low-Bit Quantization#Mixed-Precision Quantization#Sensitivity Metric#Quantization Scale Initialization#Accuracy Preservation
댓글 수 로딩 중

[논문리뷰] Mitigating Catastrophic Forgetting in Target Language Adaptation of LLMs via Source-Shielded Updates

Nikolaos Aletras이 arXiv에 게시한 'Mitigating Catastrophic Forgetting in Target Language Adaptation of LLMs via Source-Shielded Updates' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models (LLMs)#Catastrophic Forgetting#Language Adaptation#Continual Pre-training#Parameter Freezing#Low-Resource Languages#Source Knowledge Preservation
댓글 수 로딩 중

[논문리뷰] Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models

arXiv에 게시된 'Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Large Language Models (MLLMs)#Token Pruning#Graph-Structured Pruning (GSP)#Query-Conditioned Semantic Pruning (QCSP)#Determinantal Point Processes (DPP)#Model Efficiency#Visual Redundancy
댓글 수 로딩 중

[논문리뷰] Infinity-RoPE: Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout

Pinar Yanardag이 arXiv에 게시한 'Infinity-RoPE: Action-Controllable Infinite Video Generation Emerges From Autoregressive Self-Rollout' 논문에 대한 자세한 리뷰입니다.
#Review#Autoregressive Video Generation#Rotary Positional Embedding#Infinite Video Generation#Action Control#Cinematic Transitions#Video Diffusion Models#KV Cache
댓글 수 로딩 중

[논문리뷰] UltraFlux: Data-Model Co-Design for High-quality Native 4K Text-to-Image Generation across Diverse Aspect Ratios

arXiv에 게시된 'UltraFlux: Data-Model Co-Design for High-quality Native 4K Text-to-Image Generation across Diverse Aspect Ratios' 논문에 대한 자세한 리뷰입니다.
#Review#Text-to-Image Generation#Diffusion Transformers#4K Resolution#Aspect Ratio Extrapolation#Data-Model Co-Design#VAE Post-training#Positional Encoding#Diffusion Models
댓글 수 로딩 중

[논문리뷰] SyncMV4D: Synchronized Multi-view Joint Diffusion of Appearance and Motion for Hand-Object Interaction Synthesis

Hongwen Zhang이 arXiv에 게시한 'SyncMV4D: Synchronized Multi-view Joint Diffusion of Appearance and Motion for Hand-Object Interaction Synthesis' 논문에 대한 자세한 리뷰입니다.
#Review#Hand-Object Interaction#Multi-view Video Generation#4D Motion Synthesis#Diffusion Models#Spatio-temporal Consistency#Geometric Consistency#Appearance and Motion Joint Modeling
댓글 수 로딩 중

[논문리뷰] VLA-4D: Embedding 4D Awareness into Vision-Language-Action Models for SpatioTemporally Coherent Robotic Manipulation

Gim Hee Lee이 arXiv에 게시한 'VLA-4D: Embedding 4D Awareness into Vision-Language-Action Models for SpatioTemporally Coherent Robotic Manipulation' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language-Action Models#Robotic Manipulation#SpatioTemporal Coherence#4D Awareness#Visual Representation#Action Representation#Cross-Attention
댓글 수 로딩 중

[논문리뷰] Downscaling Intelligence: Exploring Perception and Reasoning Bottlenecks in Small Multimodal Models

Serena Yeung-Levy이 arXiv에 게시한 'Downscaling Intelligence: Exploring Perception and Reasoning Bottlenecks in Small Multimodal Models' 논문에 대한 자세한 리뷰입니다.
#Review#Small Multimodal Models#LLM Downscaling#Perception Bottleneck#Reasoning Bottleneck#Visual Extraction Tuning#Chain-of-Thought Reasoning#Multimodal Learning
댓글 수 로딩 중

[논문리뷰] REVISOR: Beyond Textual Reflection, Towards Multimodal Introspective Reasoning in Long-Form Video Understanding

Jingyang Chen이 arXiv에 게시한 'REVISOR: Beyond Textual Reflection, Towards Multimodal Introspective Reasoning in Long-Form Video Understanding' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Reasoning#Long-Form Video Understanding#Self-Reflection#Reinforcement Learning#Tool-Augmented MLLMs#Visual Rethinking#Video Question Answering#Causal Attribution
댓글 수 로딩 중

[논문리뷰] AraLingBench A Human-Annotated Benchmark for Evaluating Arabic Linguistic Capabilities of Large Language Models

arXiv에 게시된 'AraLingBench A Human-Annotated Benchmark for Evaluating Arabic Linguistic Capabilities of Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Arabic LLMs#Linguistic Benchmark#Human Annotation#Natural Language Understanding#Grammar Evaluation#Morphology Analysis#Syntax Assessment#Reading Comprehension
댓글 수 로딩 중

[논문리뷰] A Brain Wave Encodes a Thousand Tokens: Modeling Inter-Cortical Neural Interactions for Effective EEG-based Emotion Recognition

G. Maragatham이 arXiv에 게시한 'A Brain Wave Encodes a Thousand Tokens: Modeling Inter-Cortical Neural Interactions for Effective EEG-based Emotion Recognition' 논문에 대한 자세한 리뷰입니다.
#Review#EEG#Emotion Recognition#Transformer Architecture#Inter-Cortical Neural Interactions#Multi-Head Attention#Brain-Computer Interface#Affective Computing
댓글 수 로딩 중

[논문리뷰] MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scaling

cyyang822이 arXiv에 게시한 'MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scaling' 논문에 대한 자세한 리뷰입니다.
#Review#Research Agent#Tool-Augmented Reasoning#Interaction Scaling#Large Language Models#Reinforcement Learning#Context Management#Open-Source AI
댓글 수 로딩 중

[논문리뷰] MicroVQA++: High-Quality Microscopy Reasoning Dataset with Weakly Supervised Graphs for Multimodal Large Language Model

Bo Yan이 arXiv에 게시한 'MicroVQA++: High-Quality Microscopy Reasoning Dataset with Weakly Supervised Graphs for Multimodal Large Language Model' 논문에 대한 자세한 리뷰입니다.
#Review#Microscopy VQA#Multimodal LLM#Weak Supervision#Graph Neural Networks#Dataset Generation#Biomedical Imaging#Scientific Reasoning#Cross-Modal Consistency
댓글 수 로딩 중

[논문리뷰] CATS-V2V: A Real-World Vehicle-to-Vehicle Cooperative Perception Dataset with Complex Adverse Traffic Scenarios

Juyoung Oh이 arXiv에 게시한 'CATS-V2V: A Real-World Vehicle-to-Vehicle Cooperative Perception Dataset with Complex Adverse Traffic Scenarios' 논문에 대한 자세한 리뷰입니다.
#Review#Cooperative Perception#Vehicle-to-Vehicle (V2V)#Autonomous Driving#Dataset#Adverse Traffic Scenarios#Sensor Fusion#Temporal Alignment#3D Bounding Box Annotation
댓글 수 로딩 중

[논문리뷰] MuSc-V2: Zero-Shot Multimodal Industrial Anomaly Classification and Segmentation with Mutual Scoring of Unlabeled Samples

arXiv에 게시된 'MuSc-V2: Zero-Shot Multimodal Industrial Anomaly Classification and Segmentation with Mutual Scoring of Unlabeled Samples' 논문에 대한 자세한 리뷰입니다.
#Review#Zero-Shot Learning#Anomaly Detection#Anomaly Segmentation#Multimodal#Industrial Inspection#Mutual Scoring#Unsupervised Learning#Transformer
댓글 수 로딩 중

[논문리뷰] MathSE: Improving Multimodal Mathematical Reasoning via Self-Evolving Iterative Reflection and Reward-Guided Fine-Tuning

arXiv에 게시된 'MathSE: Improving Multimodal Mathematical Reasoning via Self-Evolving Iterative Reflection and Reward-Guided Fine-Tuning' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Reasoning#Mathematical Problem Solving#Self-Evolving#Iterative Fine-Tuning#Reward Models#Reflection#Large Language Models (LLMs)
댓글 수 로딩 중

[논문리뷰] Tiny Model, Big Logic: Diversity-Driven Optimization Elicits Large-Model Reasoning Ability in VibeThinker-1.5B

arXiv에 게시된 'Tiny Model, Big Logic: Diversity-Driven Optimization Elicits Large-Model Reasoning Ability in VibeThinker-1.5B' 논문에 대한 자세한 리뷰입니다.
#Review#Small Language Models#Reasoning#Diversity Optimization#Supervised Fine-Tuning (SFT)#Reinforcement Learning (RL)#Spectrum-to-Signal Principle (SSP)#Mathematical Reasoning#Code Generation
댓글 수 로딩 중

[논문리뷰] TimeSearch-R: Adaptive Temporal Search for Long-Form Video Understanding via Self-Verification Reinforcement Learning

arXiv에 게시된 'TimeSearch-R: Adaptive Temporal Search for Long-Form Video Understanding via Self-Verification Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Long-form Video Understanding#Temporal Search#Reinforcement Learning#Self-Verification#Video-Language Models#Adaptive Search#Interleaved Reasoning
댓글 수 로딩 중

[논문리뷰] Reasoning with Confidence: Efficient Verification of LLM Reasoning Steps via Uncertainty Heads

Jiaheng Zhang이 arXiv에 게시한 'Reasoning with Confidence: Efficient Verification of LLM Reasoning Steps via Uncertainty Heads' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Reasoning Verification#Uncertainty Quantification (UQ)#UHeads#Process Reward Models (PRMs)#Chain-of-Thought (CoT)#Self-Supervised Learning#Computational Efficiency#Domain Generalization
댓글 수 로딩 중

[논문리뷰] DRIVE: Data Curation Best Practices for Reinforcement Learning with Verifiable Reward in Competitive Code Generation

arXiv에 게시된 'DRIVE: Data Curation Best Practices for Reinforcement Learning with Verifiable Reward in Competitive Code Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning with Verifiable Reward#Competitive Programming#Code Generation#Data Curation#Curriculum Learning#Supervised Fine-tuning#Entropy Expansion
댓글 수 로딩 중

[논문리뷰] When Modalities Conflict: How Unimodal Reasoning Uncertainty Governs Preference Dynamics in MLLMs

Haotian Wang이 arXiv에 게시한 'When Modalities Conflict: How Unimodal Reasoning Uncertainty Governs Preference Dynamics in MLLMs' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Large Language Models (MLLMs)#Modality Following#Unimodal Uncertainty#Modality Preference#Conflict Resolution#Internal Mechanism#Entropy#Controllable Dataset
댓글 수 로딩 중

[논문리뷰] ChartM^3: A Multi-Stage Code-Driven Pipeline for Constructing Multi-Dimensional and Multi-Step Visual Reasoning Data in Chart Comprehension

Hao Wang이 arXiv에 게시한 'ChartM^3: A Multi-Stage Code-Driven Pipeline for Constructing Multi-Dimensional and Multi-Step Visual Reasoning Data in Chart Comprehension' 논문에 대한 자세한 리뷰입니다.
#Review#Chart Comprehension#Visual Reasoning#Data Generation#Code-Driven Pipeline#Multimodal LLMs#Retrieval-Augmented Generation#Reinforcement Learning#Synthetic Data
댓글 수 로딩 중

[논문리뷰] left|,circlearrowright,text{BUS},right|: A Large and Diverse Multimodal Benchmark for evaluating the ability of Vision-Language Models to understand Rebus Puzzles

Deepiha S이 arXiv에 게시한 'left|,circlearrowright,text{BUS},right|: A Large and Diverse Multimodal Benchmark for evaluating the ability of Vision-Language Models to understand Rebus Puzzles' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language Models#Multimodal Benchmark#Rebus Puzzles#In-Context Learning#Reasoning#ControlNet#Prompt Engineering
댓글 수 로딩 중

[논문리뷰] How Far Are Surgeons from Surgical World Models? A Pilot Study on Zero-shot Surgical Video Generation with Expert Assessment

Yuhao Zhai이 arXiv에 게시한 'How Far Are Surgeons from Surgical World Models? A Pilot Study on Zero-shot Surgical Video Generation with Expert Assessment' 논문에 대한 자세한 리뷰입니다.
#Review#Video Generation#World Models#Surgical AI#Zero-shot Prediction#Expert Evaluation#Plausibility Gap#Medical Simulation
댓글 수 로딩 중

[논문리뷰] MisSynth: Improving MISSCI Logical Fallacies Classification with Synthetic Data

Nadiya Shvai이 arXiv에 게시한 'MisSynth: Improving MISSCI Logical Fallacies Classification with Synthetic Data' 논문에 대한 자세한 리뷰입니다.
#Review#Health Misinformation#Logical Fallacy Classification#Synthetic Data Generation#Large Language Models (LLMs)#Retrieval-Augmented Generation (RAG)#Parameter-Efficient Fine-tuning (PEFT)#LoRA#MISSCI Benchmark
댓글 수 로딩 중

[논문리뷰] Mask-to-Height: A YOLOv11-Based Architecture for Joint Building Instance Segmentation and Height Classification from Satellite Imagery

Oğuz Hanoğlu이 arXiv에 게시한 'Mask-to-Height: A YOLOv11-Based Architecture for Joint Building Instance Segmentation and Height Classification from Satellite Imagery' 논문에 대한 자세한 리뷰입니다.
#Review#Building Instance Segmentation#Height Classification#YOLOv11#Satellite Imagery#Multitask Learning#Remote Sensing#Urban Planning
댓글 수 로딩 중

[논문리뷰] Limits of Generalization in RLVR: Two Case Studies in Mathematical Reasoning

Nidhi Rastogi이 arXiv에 게시한 'Limits of Generalization in RLVR: Two Case Studies in Mathematical Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning with Verifiable Rewards (RLVR)#Mathematical Reasoning#Large Language Models (LLMs)#Activity Scheduling#Longest Increasing Subsequence (LIS)#Generalization Limits#Reward Design#Self-consistency
댓글 수 로딩 중

[논문리뷰] MedVLSynther: Synthesizing High-Quality Visual Question Answering from Medical Documents with Generator-Verifier LMMs

arXiv에 게시된 'MedVLSynther: Synthesizing High-Quality Visual Question Answering from Medical Documents with Generator-Verifier LMMs' 논문에 대한 자세한 리뷰입니다.
#Review#Medical VQA#Large Multimodal Models (LMMs)#Data Synthesis#Generator-Verifier Framework#Rubric-Guided#Reinforcement Learning (RL)#Context-Aware
댓글 수 로딩 중

[논문리뷰] The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution

Haoze Wu이 arXiv에 게시한 'The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution' 논문에 대한 자세한 리뷰입니다.
#Review#Language Agents#Tool Use#Benchmarking#Long-Horizon Tasks#Realistic Environments#Multi-Application#Execution-Based Evaluation#Model Context Protocol (MCP)
댓글 수 로딩 중

[논문리뷰] LightBagel: A Light-weighted, Double Fusion Framework for Unified Multimodal Understanding and Generation

Chaorui Deng이 arXiv에 게시한 'LightBagel: A Light-weighted, Double Fusion Framework for Unified Multimodal Understanding and Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Unified Multimodal Models#Double Fusion#Lightweight AI#Text-to-Image Generation#Image Editing#Model Architecture#Efficient Training#Cross-modal Interaction
댓글 수 로딩 중

[논문리뷰] ALICE-LRI: A General Method for Lossless Range Image Generation for Spinning LiDAR Sensors without Calibration Metadata

José C. Cabaleiro이 arXiv에 게시한 'ALICE-LRI: A General Method for Lossless Range Image Generation for Spinning LiDAR Sensors without Calibration Metadata' 논문에 대한 자세한 리뷰입니다.
#Review#LiDAR#Range Image#Lossless Projection#Sensor Calibration#Intrinsic Parameters#Point Cloud Reconstruction#Hough Transform#Weighted Least Squares
댓글 수 로딩 중

[논문리뷰] ComProScanner: A multi-agent based framework for composition-property structured data extraction from scientific literature

arXiv에 게시된 'ComProScanner: A multi-agent based framework for composition-property structured data extraction from scientific literature' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-agent Systems#Large Language Models (LLMs)#Information Extraction#Scientific Literature#Materials Science#Data Curation#Piezoelectric Materials#RAG (Retrieval-Augmented Generation)
댓글 수 로딩 중

[논문리뷰] KORE: Enhancing Knowledge Injection for Large Multimodal Models via Knowledge-Oriented Augmentations and Constraints

Jinhe Bi이 arXiv에 게시한 'KORE: Enhancing Knowledge Injection for Large Multimodal Models via Knowledge-Oriented Augmentations and Constraints' 논문에 대한 자세한 리뷰입니다.
#Review#Knowledge Injection#Large Multimodal Models#Catastrophic Forgetting#Data Augmentation#Parameter-Efficient Fine-Tuning#Null Space#Continual Learning
댓글 수 로딩 중

[논문리뷰] BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping

Junrui Shen이 arXiv에 게시한 'BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping' 논문에 대한 자세한 리뷰입니다.
#Review#Off-Policy Reinforcement Learning#Large Language Models#Adaptive Clipping#Policy Optimization#PPO#Entropy Preservation#RL Stabilization
댓글 수 로딩 중

[논문리뷰] AlphaOPT: Formulating Optimization Programs with Self-Improving LLM Experience Library

Chonghe Jiang이 arXiv에 게시한 'AlphaOPT: Formulating Optimization Programs with Self-Improving LLM Experience Library' 논문에 대한 자세한 리뷰입니다.
#Review#Optimization Modeling#Large Language Models (LLMs)#Experience Library#Self-Improving Systems#Continual Learning#Out-of-Distribution Generalization#Operations Research#Knowledge Representation
댓글 수 로딩 중

[논문리뷰] Unleashing Scientific Reasoning for Bio-experimental Protocol Generation via Structured Component-based Reward Mechanism

Shuang Gu이 arXiv에 게시한 'Unleashing Scientific Reasoning for Bio-experimental Protocol Generation via Structured Component-based Reward Mechanism' 논문에 대한 자세한 리뷰입니다.
#Review#Scientific Reasoning#Bio-experimental Protocol Generation#LLM#Structured Reward#SciRecipe Dataset#Sketch-and-Fill#Reinforcement Learning#Thoth
댓글 수 로딩 중

[논문리뷰] PokeeResearch: Effective Deep Research via Reinforcement Learning from AI Feedback and Robust Reasoning Scaffold

arXiv에 게시된 'PokeeResearch: Effective Deep Research via Reinforcement Learning from AI Feedback and Robust Reasoning Scaffold' 논문에 대한 자세한 리뷰입니다.
#Review#Deep Research Agent#Reinforcement Learning from AI Feedback#RLOO Algorithm#Large Language Models#Tool Use#Self-Correction#Reasoning Scaffold#Agent Alignment
댓글 수 로딩 중

[논문리뷰] Balanced Multi-Task Attention for Satellite Image Classification: A Systematic Approach to Achieving 97.23% Accuracy on EuroSAT Without Pre-Training

Aditya Vir이 arXiv에 게시한 'Balanced Multi-Task Attention for Satellite Image Classification: A Systematic Approach to Achieving 97.23% Accuracy on EuroSAT Without Pre-Training' 논문에 대한 자세한 리뷰입니다.
#Review#Satellite Image Classification#Multi-Task Attention#From-Scratch Training#EuroSAT Dataset#Squeeze-Excitation Networks#Coordinate Attention#CNN#Deep Learning Architecture
댓글 수 로딩 중

[논문리뷰] Rewiring Experts on the Fly:Continuous Rerouting for Better Online Adaptation in Mixture-of-Expert models

Shiwei Liu이 arXiv에 게시한 'Rewiring Experts on the Fly:Continuous Rerouting for Better Online Adaptation in Mixture-of-Expert models' 논문에 대한 자세한 리뷰입니다.
#Review#Mixture-of-Experts (MoE)#Online Adaptation#Test-Time Adaptation (TTA)#Expert Routing#Large Language Models (LLMs)#Self-Supervision#Computational Efficiency#Context Shift Robustness
댓글 수 로딩 중

[논문리뷰] ERGO: Entropy-guided Resetting for Generation Optimization in Multi-turn Language Models

Sean O'Brien이 arXiv에 게시한 'ERGO: Entropy-guided Resetting for Generation Optimization in Multi-turn Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-turn Conversation#Large Language Models (LLMs)#Context Management#Entropy-guided Resetting#Uncertainty Quantification#Performance Degradation#Prompt Engineering#Conversational AI
댓글 수 로딩 중

[논문리뷰] Build Your Personalized Research Group: A Multiagent Framework for Continual and Interactive Science Automation

Cat Yan이 arXiv에 게시한 'Build Your Personalized Research Group: A Multiagent Framework for Continual and Interactive Science Automation' 논문에 대한 자세한 리뷰입니다.
#Review#Multiagent Systems#Science Automation#Dynamic Workflows#Workspace-based Communication#Context Compaction#Human-in-the-loop AI#Open-source Framework
댓글 수 로딩 중

[논문리뷰] VLA^2: Empowering Vision-Language-Action Models with an Agentic Framework for Unseen Concept Manipulation

arXiv에 게시된 'VLA^2: Empowering Vision-Language-Action Models with an Agentic Framework for Unseen Concept Manipulation' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language-Action Models#Agentic Framework#Unseen Concept Manipulation#Out-of-Distribution Generalization#Tool Use#Web Retrieval#Object Detection#LIBERO Simulation
댓글 수 로딩 중

[논문리뷰] MoM: Mixtures of Scenario-Aware Document Memories for Retrieval-Augmented Generation Systems

Feiyu Xiong이 arXiv에 게시한 'MoM: Mixtures of Scenario-Aware Document Memories for Retrieval-Augmented Generation Systems' 논문에 대한 자세한 리뷰입니다.
#Review#Retrieval-Augmented Generation (RAG)#Document Memory#Text Chunking#Small Language Models (SLMs)#Large Language Models (LLMs)#Scenario-Aware Processing#Multi-Layer Retrieval#Cognitive Simulation
댓글 수 로딩 중

[논문리뷰] Hierarchical Frequency Tagging Probe (HFTP): A Unified Approach to Investigate Syntactic Structure Representations in Large Language Models and the Human Brain

Lingxi Lu이 arXiv에 게시한 'Hierarchical Frequency Tagging Probe (HFTP): A Unified Approach to Investigate Syntactic Structure Representations in Large Language Models and the Human Brain' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models#Syntactic Structure#Human Brain#Frequency Tagging#Neuroscience#Model Interpretability#Representational Similarity Analysis#Intracranial EEG
댓글 수 로딩 중

[논문리뷰] CVD-STORM: Cross-View Video Diffusion with Spatial-Temporal Reconstruction Model for Autonomous Driving

Jingcheng Ni이 arXiv에 게시한 'CVD-STORM: Cross-View Video Diffusion with Spatial-Temporal Reconstruction Model for Autonomous Driving' 논문에 대한 자세한 리뷰입니다.
#Review#Autonomous Driving#Video Generation#Diffusion Models#Spatial-Temporal Reconstruction#3D Gaussian Splatting#Variational Autoencoder#World Modeling#Multi-View Video
댓글 수 로딩 중

[논문리뷰] ReFIne: A Framework for Trustworthy Large Reasoning Models with Reliability, Faithfulness, and Interpretability

Tsui-Wei Weng이 arXiv에 게시한 'ReFIne: A Framework for Trustworthy Large Reasoning Models with Reliability, Faithfulness, and Interpretability' 논문에 대한 자세한 리뷰입니다.
#Review#Trustworthy AI#Large Reasoning Models (LRMs)#Interpretability#Faithfulness#Reliability#Chain-of-Thought (CoT)#Supervised Fine-tuning (SFT)#GRPO
댓글 수 로딩 중

[논문리뷰] Hybrid-grained Feature Aggregation with Coarse-to-fine Language Guidance for Self-supervised Monocular Depth Estimation

Zekun Qi이 arXiv에 게시한 'Hybrid-grained Feature Aggregation with Coarse-to-fine Language Guidance for Self-supervised Monocular Depth Estimation' 논문에 대한 자세한 리뷰입니다.
#Review#Self-supervised Monocular Depth Estimation#Foundation Models#CLIP#DINO#Language Guidance#Coarse-to-fine Learning#Feature Aggregation#3D Perception
댓글 수 로딩 중

[논문리뷰] Bridging Reasoning to Learning: Unmasking Illusions using Complexity Out of Distribution Generalization

Mahdi Ghaznavai이 arXiv에 게시한 'Bridging Reasoning to Learning: Unmasking Illusions using Complexity Out of Distribution Generalization' 논문에 대한 자세한 리뷰입니다.
#Review#Complexity OoD Generalization#System-1 Thinking#System-2 Reasoning#Kolmogorov Complexity#Inductive Biases#Large Language Models (LLMs)#Reasoning Evaluation
댓글 수 로딩 중

[논문리뷰] A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks

Fanchao Qi이 arXiv에 게시한 'A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks' 논문에 대한 자세한 리뷰입니다.
#Review#Long-Horizon Tasks#LLM Agents#Global Planning#Reinforcement Learning#Supervised Fine-tuning#Homologous Consensus Filtering#Executor Capability Gain Reward#Plan-and-Execute
댓글 수 로딩 중

[논문리뷰] Recycling Pretrained Checkpoints: Orthogonal Growth of Mixture-of-Experts for Efficient Large Language Model Pre-Training

Peng Cheng이 arXiv에 게시한 'Recycling Pretrained Checkpoints: Orthogonal Growth of Mixture-of-Experts for Efficient Large Language Model Pre-Training' 논문에 대한 자세한 리뷰입니다.
#Review#Mixture-of-Experts#Large Language Models#Checkpoint Recycling#Model Growth#Efficient Pretraining#Depth Growth#Width Growth#Sunk Cost
댓글 수 로딩 중

[논문리뷰] MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization

vanilla1116이 arXiv에 게시한 'MM-HELIX: Boosting Multimodal Long-Chain Reflective Reasoning with Holistic Platform and Adaptive Hybrid Policy Optimization' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal LLMs#Reflective Reasoning#Long-Chain Reasoning#Benchmark#Policy Optimization#Data Generation#Reinforcement Learning#Backtracking
댓글 수 로딩 중

[논문리뷰] Entropy Regularizing Activation: Boosting Continuous Control, Large Language Models, and Image Classification with Activation as Entropy Constraints

Huazhe Xu이 arXiv에 게시한 'Entropy Regularizing Activation: Boosting Continuous Control, Large Language Models, and Image Classification with Activation as Entropy Constraints' 논문에 대한 자세한 리뷰입니다.
#Review#Entropy Regularization#Activation Functions#Continuous Control#Large Language Models#Image Classification#Reinforcement Learning#Policy Stochasticity#Entropy Constraints
댓글 수 로딩 중

[논문리뷰] Patch-as-Decodable-Token: Towards Unified Multi-Modal Vision Tasks in MLLMs

Jingyi Liao이 arXiv에 게시한 'Patch-as-Decodable-Token: Towards Unified Multi-Modal Vision Tasks in MLLMs' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Large Language Models (MLLMs)#Visual Reference Tokens (VRTs)#Dense Prediction#Referring Expression Comprehension (REC)#Open-Vocabulary Detection (OVD)#Image Captioning#Unified Architecture#Autoregressive Generation
댓글 수 로딩 중

[논문리뷰] D^3QE: Learning Discrete Distribution Discrepancy-aware Quantization Error for Autoregressive-Generated Image Detection

Yueqi Duan이 arXiv에 게시한 'D^3QE: Learning Discrete Distribution Discrepancy-aware Quantization Error for Autoregressive-Generated Image Detection' 논문에 대한 자세한 리뷰입니다.
#Review#Autoregressive Models#Image Detection#Discrete Distribution Discrepancy#Quantization Error#Transformer#Generative AI#Deepfake Detection
댓글 수 로딩 중

[논문리뷰] AlphaApollo: Orchestrating Foundation Models and Professional Tools into a Self-Evolving System for Deep Agentic Reasoning

Zongze Li이 arXiv에 게시한 'AlphaApollo: Orchestrating Foundation Models and Professional Tools into a Self-Evolving System for Deep Agentic Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Foundation Models#Agentic Reasoning#Tool Use#Self-Evolving System#Retrieval-Augmented Generation#Computational Tools#Error Correction
댓글 수 로딩 중

[논문리뷰] Revisiting Modeling and Evaluation Approaches in Speech Emotion Recognition: Considering Subjectivity of Annotators and Ambiguity of Emotions

arXiv에 게시된 'Revisiting Modeling and Evaluation Approaches in Speech Emotion Recognition: Considering Subjectivity of Annotators and Ambiguity of Emotions' 논문에 대한 자세한 리뷰입니다.
#Review#Speech Emotion Recognition#Annotator Subjectivity#Emotion Ambiguity#Soft Labels#Multi-label Classification#Evaluation Metrics#Loss Functions
댓글 수 로딩 중

[논문리뷰] HalluGuard: Evidence-Grounded Small Reasoning Models to Mitigate Hallucinations in Retrieval-Augmented Generation

Radu State이 arXiv에 게시한 'HalluGuard: Evidence-Grounded Small Reasoning Models to Mitigate Hallucinations in Retrieval-Augmented Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Hallucination Detection#Retrieval-Augmented Generation (RAG)#Small Reasoning Model (SRM)#Preference Fine-tuning#ORPO#Evidence Grounding#Fact-checking
댓글 수 로딩 중

[논문리뷰] Distributional Semantics Tracing: A Framework for Explaining Hallucinations in Large Language Models

Jacobo Azcona이 arXiv에 게시한 'Distributional Semantics Tracing: A Framework for Explaining Hallucinations in Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Hallucinations#Mechanistic Interpretability#Distributional Semantics Tracing (DST)#Dual-Process Theory#Semantic Drift#Commitment Layer#Faithfulness Score
댓글 수 로딩 중

[논문리뷰] Benchmark It Yourself (BIY): Preparing a Dataset and Benchmarking AI Models for Scatterplot-Related Tasks

Pedro Bizarro이 arXiv에 게시한 'Benchmark It Yourself (BIY): Preparing a Dataset and Benchmarking AI Models for Scatterplot-Related Tasks' 논문에 대한 자세한 리뷰입니다.
#Review#Scatterplot Analysis#AI Benchmarking#Multimodal LLMs#Synthetic Data Generation#Cluster Detection#Outlier Detection#Data Visualization#Prompt Engineering
댓글 수 로딩 중

[논문리뷰] EvolProver: Advancing Automated Theorem Proving by Evolving Formalized Problems via Symmetry and Difficulty

Xuanwu Wang이 arXiv에 게시한 'EvolProver: Advancing Automated Theorem Proving by Evolving Formalized Problems via Symmetry and Difficulty' 논문에 대한 자세한 리뷰입니다.
#Review#Automated Theorem Proving#Data Augmentation#Large Language Models#Formal Mathematics#Symmetry#Difficulty Evolution#Abstract Syntax Tree#Generalizability
댓글 수 로딩 중

[논문리뷰] AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning

Zeliang Zhang이 arXiv에 게시한 'AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-Agent Reinforcement Learning#Adversarial Co-evolution#LLM Safety#Jailbreak Attacks#Internalized Safety#Public Baseline#System Robustness
댓글 수 로딩 중

[논문리뷰] Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned

arXiv에 게시된 'Training Vision-Language Process Reward Models for Test-Time Scaling in Multimodal Reasoning: Key Insights and Lessons Learned' 논문에 대한 자세한 리뷰입니다.
#Review#Vision-Language Models (VLMs)#Process Reward Models (PRMs)#Multimodal Reasoning#Test-Time Scaling (TTS)#Process Supervision#Dataset Construction#Perception Errors#MCTS
댓글 수 로딩 중

[논문리뷰] DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search

arXiv에 게시된 'DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning with Verifiable Rewards (RLVR)#Monte Carlo Tree Search (MCTS)#Mathematical Reasoning#Large Language Models (LLMs)#Systematic Exploration#Adaptive Training#Tree-GRPO
댓글 수 로딩 중

[논문리뷰] Beyond Log Likelihood: Probability-Based Objectives for Supervised Fine-Tuning across the Model Capability Continuum

Hanghang Tong이 arXiv에 게시한 'Beyond Log Likelihood: Probability-Based Objectives for Supervised Fine-Tuning across the Model Capability Continuum' 논문에 대한 자세한 리뷰입니다.
#Review#Supervised Fine-tuning (SFT)#Large Language Models (LLMs)#Training Objectives#Negative Log Likelihood (NLL)#Model Capability Continuum#Generalization#Probability-based Loss Functions
댓글 수 로딩 중

[논문리뷰] Specialization after Generalization: Towards Understanding Test-Time Training in Foundation Models

arXiv에 게시된 'Specialization after Generalization: Towards Understanding Test-Time Training in Foundation Models' 논문에 대한 자세한 리뷰입니다.
#Review#Test-Time Training (TTT)#Foundation Models#Underparameterization#Sparse Autoencoders (SAE)#Linear Representation Hypothesis (LRH)#Specialization#Scaling Laws#In-Distribution Data
댓글 수 로딩 중

[논문리뷰] A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects

Jennifer Ding이 arXiv에 게시한 'A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects' 논문에 대한 자세한 리뷰입니다.
#Review#Open Source AI#LLM Development#Open Collaboration#Governance Models#Developer Motivations#Community Engagement#AI Ecosystem
댓글 수 로딩 중

[논문리뷰] Think-on-Graph 3.0: Efficient and Adaptive LLM Reasoning on Heterogeneous Graphs via Multi-Agent Dual-Evolving Context Retrieval

arXiv에 게시된 'Think-on-Graph 3.0: Efficient and Adaptive LLM Reasoning on Heterogeneous Graphs via Multi-Agent Dual-Evolving Context Retrieval' 논문에 대한 자세한 리뷰입니다.
#Review#RAG#LLM Reasoning#Knowledge Graphs#Multi-Agent Systems#Context Retrieval#Heterogeneous Graphs#Adaptive Learning#Dual-Evolution
댓글 수 로딩 중

[논문리뷰] No Prompt Left Behind: Exploiting Zero-Variance Prompts in LLM Reinforcement Learning via Entropy-Guided Advantage Shaping

arXiv에 게시된 'No Prompt Left Behind: Exploiting Zero-Variance Prompts in LLM Reinforcement Learning via Entropy-Guided Advantage Shaping' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Reinforcement Learning#Zero-Variance Prompts#Advantage Shaping#Entropy-Guided#Math Reasoning#RLVR#Group Relative Policy Optimization
댓글 수 로딩 중

[논문리뷰] Mind-the-Glitch: Visual Correspondence for Detecting Inconsistencies in Subject-Driven Generation

Peter Wonka이 arXiv에 게시한 'Mind-the-Glitch: Visual Correspondence for Detecting Inconsistencies in Subject-Driven Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Subject-Driven Generation#Visual Inconsistency Detection#Feature Disentanglement#Diffusion Models#Semantic Correspondence#Evaluation Metric#Spatial Localization#Contrastive Learning
댓글 수 로딩 중

[논문리뷰] Learn the Ropes, Then Trust the Wins: Self-imitation with Progressive Exploration for Agentic Reinforcement Learning

Gang Li이 arXiv에 게시한 'Learn the Ropes, Then Trust the Wins: Self-imitation with Progressive Exploration for Agentic Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#LLM Agents#Exploration-Exploitation#Self-Imitation Learning#Intrinsic Rewards#Curriculum Learning#Policy Entropy#Tool Use
댓글 수 로딩 중

[논문리뷰] CHURRO: Making History Readable with an Open-Weight Large Vision-Language Model for High-Accuracy, Low-Cost Historical Text Recognition

arXiv에 게시된 'CHURRO: Making History Readable with an Open-Weight Large Vision-Language Model for High-Accuracy, Low-Cost Historical Text Recognition' 논문에 대한 자세한 리뷰입니다.
#Review#Historical Text Recognition#Vision-Language Model#Open-Weight Model#OCR#Cultural Heritage#Low-Cost AI#Dataset Curation#Fine-tuning
댓글 수 로딩 중

[논문리뷰] Recon-Act: A Self-Evolving Multi-Agent Browser-Use System via Web Reconnaissance, Tool Generation, and Task Execution

Jinjie Gu이 arXiv에 게시한 'Recon-Act: A Self-Evolving Multi-Agent Browser-Use System via Web Reconnaissance, Tool Generation, and Task Execution' 논문에 대한 자세한 리뷰입니다.
#Review#Multi-Agent System#Browser Automation#Web Reconnaissance#Tool Generation#Task Execution#Self-Evolving AI#LLM/VLM#VisualWebArena
댓글 수 로딩 중

[논문리뷰] MI-Fuse: Label Fusion for Unsupervised Domain Adaptation with Closed-Source Large-Audio Language Model

Hung-yi Lee이 arXiv에 게시한 'MI-Fuse: Label Fusion for Unsupervised Domain Adaptation with Closed-Source Large-Audio Language Model' 논문에 대한 자세한 리뷰입니다.
#Review#Speech Emotion Recognition#Source-Free Unsupervised Domain Adaptation#Large Audio-Language Models#Label Fusion#Mutual Information#API-Only Models#Domain Mismatch
댓글 수 로딩 중

[논문리뷰] Zero-Shot Multi-Spectral Learning: Reimagining a Generalist Multimodal Gemini 2.5 Model for Remote Sensing Applications

Genady Beryozkin이 arXiv에 게시한 'Zero-Shot Multi-Spectral Learning: Reimagining a Generalist Multimodal Gemini 2.5 Model for Remote Sensing Applications' 논문에 대한 자세한 리뷰입니다.
#Review#Remote Sensing#Zero-Shot Learning#Multimodal Models#Multi-spectral Imagery#Gemini 2.5#Prompt Engineering#Land Cover Classification#Pseudo-Image
댓글 수 로딩 중

[논문리뷰] When Big Models Train Small Ones: Label-Free Model Parity Alignment for Efficient Visual Question Answering using Small VLMs

Anand Mishra이 arXiv에 게시한 'When Big Models Train Small Ones: Label-Free Model Parity Alignment for Efficient Visual Question Answering using Small VLMs' 논문에 대한 자세한 리뷰입니다.
#Review#VQA#Small VLMs#Large VLMs#Knowledge Transfer#Pseudo-labeling#Label-Free Learning#Model Parity Alignment#Computational Efficiency
댓글 수 로딩 중

[논문리뷰] FlagEval Findings Report: A Preliminary Evaluation of Large Reasoning Models on Automatically Verifiable Textual and Visual Questions

tengdai722이 arXiv에 게시한 'FlagEval Findings Report: A Preliminary Evaluation of Large Reasoning Models on Automatically Verifiable Textual and Visual Questions' 논문에 대한 자세한 리뷰입니다.
#Review#Large Reasoning Models#LLM Evaluation#Multimodal AI#Reasoning Behaviors#Hallucination#Contamination-Free#AI Safety#Instruction Following
댓글 수 로딩 중

[논문리뷰] DIWALI - Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context

Maunendra Sankar Desarkar이 arXiv에 게시한 'DIWALI - Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context' 논문에 대한 자세한 리뷰입니다.
#Review#Cultural Adaptation#Large Language Models#Indian Culture#Dataset Creation#CSI#Human Evaluation#LLM Evaluation#Cultural Bias
댓글 수 로딩 중

[논문리뷰] Analyzing the Effects of Supervised Fine-Tuning on Model Knowledge from Token and Parameter Levels

Qi Zhang이 arXiv에 게시한 'Analyzing the Effects of Supervised Fine-Tuning on Model Knowledge from Token and Parameter Levels' 논문에 대한 자세한 리뷰입니다.
#Review#Supervised Fine-Tuning (SFT)#Large Language Models (LLMs)#Model Knowledge#Closed-Book Question Answering (CBQA)#Parameter Restoration#Kullback-Leibler Divergence#Knowledge Forgetting
댓글 수 로딩 중

[논문리뷰] Do You Hear What I Mean? Quantifying the Instruction-Perception Gap in Instruction-Guided Expressive Text-To-Speech Systems

Hung-yi Lee이 arXiv에 게시한 'Do You Hear What I Mean? Quantifying the Instruction-Perception Gap in Instruction-Guided Expressive Text-To-Speech Systems' 논문에 대한 자세한 리뷰입니다.
#Review#Instruction-Guided TTS#Expressive Speech Synthesis#Human Perception#Subjective Evaluation#Controllability#Instruction Following#Evaluation Metrics
댓글 수 로딩 중

[논문리뷰] Unleashing the Potential of Multimodal LLMs for Zero-Shot Spatio-Temporal Video Grounding

Rynson W. H. Lau이 arXiv에 게시한 'Unleashing the Potential of Multimodal LLMs for Zero-Shot Spatio-Temporal Video Grounding' 논문에 대한 자세한 리뷰입니다.
#Review#Spatio-Temporal Video Grounding#Multimodal Large Language Models#Zero-Shot Learning#Visual Grounding#Decomposed Spatio-Temporal Highlighting#Logit-Guided Re-attention#Temporal-Augmented Assembling
댓글 수 로딩 중

[논문리뷰] MARS2 2025 Challenge on Multimodal Reasoning: Datasets, Methods, Results, Discussion, and Outlook

Bowen Zhou이 arXiv에 게시한 'MARS2 2025 Challenge on Multimodal Reasoning: Datasets, Methods, Results, Discussion, and Outlook' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Reasoning#Large Language Models (LLMs)#Multimodal Large Language Models (MLLMs)#Visual Grounding#Visual Question Answering#Advertisement Video Analysis#Real-world Scenarios#Challenge Benchmark
댓글 수 로딩 중

[논문리뷰] Multiple Instance Learning Framework with Masked Hard Instance Mining for Gigapixel Histopathology Image Analysis

Bo Liu이 arXiv에 게시한 'Multiple Instance Learning Framework with Masked Hard Instance Mining for Gigapixel Histopathology Image Analysis' 논문에 대한 자세한 리뷰입니다.
#Review#Multiple Instance Learning#Hard Instance Mining#Computational Pathology#Whole Slide Images#Masked Learning#Siamese Network#Medical Image Analysis
댓글 수 로딩 중

[논문리뷰] Multimodal Reasoning for Science: Technical Report and 1st Place Solution to the ICML 2025 SeePhys Challenge

Wentao Zhang이 arXiv에 게시한 'Multimodal Reasoning for Science: Technical Report and 1st Place Solution to the ICML 2025 SeePhys Challenge' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Reasoning#Science AI#Caption-assisted Reasoning#SeePhys Challenge#Large Language Models#Visual Question Answering#Physics Problems#Cross-modal Alignment
댓글 수 로딩 중

[논문리뷰] LazyDrag: Enabling Stable Drag-Based Editing on Multi-Modal Diffusion Transformers via Explicit Correspondence

Lionel M. Ni이 arXiv에 게시한 'LazyDrag: Enabling Stable Drag-Based Editing on Multi-Modal Diffusion Transformers via Explicit Correspondence' 논문에 대한 자세한 리뷰입니다.
#Review#Image Editing#Diffusion Models#Multi-Modal Transformers#Drag-based Editing#Explicit Correspondence#Attention Control#Identity Preservation#Training-Free
댓글 수 로딩 중

[논문리뷰] Dr.V: A Hierarchical Perception-Temporal-Cognition Framework to Diagnose Video Hallucination by Fine-grained Spatial-Temporal Grounding

Li Zheng이 arXiv에 게시한 'Dr.V: A Hierarchical Perception-Temporal-Cognition Framework to Diagnose Video Hallucination by Fine-grained Spatial-Temporal Grounding' 논문에 대한 자세한 리뷰입니다.
#Review#Video Hallucination#Large Video Models (LVMs)#Hierarchical Reasoning#Spatial-Temporal Grounding#Diagnostic Framework#Benchmark Dataset#Multimodal AI
댓글 수 로딩 중

[논문리뷰] The Choice of Divergence: A Neglected Key to Mitigating Diversity Collapse in Reinforcement Learning with Verifiable Reward

Xiaoyu Tan이 arXiv에 게시한 'The Choice of Divergence: A Neglected Key to Mitigating Diversity Collapse in Reinforcement Learning with Verifiable Reward' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Large Language Models (LLMs)#Diversity Collapse#f-divergence#Forward-KL#JS-divergence#Pass@k#Catastrophic Forgetting
댓글 수 로딩 중

[논문리뷰] Kling-Avatar: Grounding Multimodal Instructions for Cascaded Long-Duration Avatar Animation Synthesis

Wentao Hu이 arXiv에 게시한 'Kling-Avatar: Grounding Multimodal Instructions for Cascaded Long-Duration Avatar Animation Synthesis' 논문에 대한 자세한 리뷰입니다.
#Review#Avatar Animation#Multimodal Instructions#Long-Duration Video Generation#MLLM Director#Cascaded Framework#Lip Synchronization#Instruction Grounding#Video Diffusion Transformers
댓글 수 로딩 중

[논문리뷰] AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning

Honglin Guo이 arXiv에 게시한 'AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Agents#Reinforcement Learning#Multi-Turn Interaction#Long-Horizon Decision Making#Agent Framework#Exploration-Exploitation#Progressive Scaling
댓글 수 로딩 중

[논문리뷰] MedVista3D: Vision-Language Modeling for Reducing Diagnostic Errors in 3D CT Disease Detection, Understanding and Reporting

Vanessa Wildman이 arXiv에 게시한 'MedVista3D: Vision-Language Modeling for Reducing Diagnostic Errors in 3D CT Disease Detection, Understanding and Reporting' 논문에 대한 자세한 리뷰입니다.
#Review#3D CT#Vision-Language Model#Medical Imaging#Diagnostic Error Reduction#Multi-scale Alignment#Semantic Enrichment#Radiology Reporting#Zero-shot Learning
댓글 수 로딩 중

[논문리뷰] LatticeWorld: A Multimodal Large Language Model-Empowered Framework for Interactive Complex World Generation

Zhan Zhao이 arXiv에 게시한 'LatticeWorld: A Multimodal Large Language Model-Empowered Framework for Interactive Complex World Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal LLM#3D World Generation#Unreal Engine 5#Procedural Content Generation#Interactive Environments#Sim-to-Real#Spatial Understanding#Multimodal Input
댓글 수 로딩 중

[논문리뷰] Towards More Diverse and Challenging Pre-training for Point Cloud Learning: Self-Supervised Cross Reconstruction with Decoupled Views

Junchi Yan이 arXiv에 게시한 'Towards More Diverse and Challenging Pre-training for Point Cloud Learning: Self-Supervised Cross Reconstruction with Decoupled Views' 논문에 대한 자세한 리뷰입니다.
#Review#Point Cloud Learning#Self-Supervised Learning#Cross Reconstruction#Decoupled Views#Generative Models#Positional Encoding#3D Vision
댓글 수 로딩 중

[논문리뷰] C-DiffDet+: Fusing Global Scene Context with Generative Denoising for High-Fidelity Object Detection

Vito Renó이 arXiv에 게시한 'C-DiffDet+: Fusing Global Scene Context with Generative Denoising for High-Fidelity Object Detection' 논문에 대한 자세한 리뷰입니다.
#Review#Object Detection#Diffusion Model#Global Scene Context#Context-Aware Fusion#Fine-grained Detection#Automotive Damage Assessment#Generative Denoising#Cross-Attention
댓글 수 로딩 중

[논문리뷰] Attributes as Textual Genes: Leveraging LLMs as Genetic Algorithm Simulators for Conditional Synthetic Data Generation

Xiaolei Huang이 arXiv에 게시한 'Attributes as Textual Genes: Leveraging LLMs as Genetic Algorithm Simulators for Conditional Synthetic Data Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Synthetic Data Generation#Large Language Models (LLMs)#Genetic Algorithms#Textual Data Augmentation#Active Learning#NLP#Data Diversity
댓글 수 로딩 중

[논문리뷰] AMBEDKAR-A Multi-level Bias Elimination through a Decoding Approach with Knowledge Augmentation for Robust Constitutional Alignment of Language Models

Rahul Karthikeyan이 arXiv에 게시한 'AMBEDKAR-A Multi-level Bias Elimination through a Decoding Approach with Knowledge Augmentation for Robust Constitutional Alignment of Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Bias Mitigation#Large Language Models#Speculative Decoding#Constitutional AI#Fairness#Inference-Time Control#Indian Sociocultural Context
댓글 수 로딩 중

[논문리뷰] R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning

Han Hu이 arXiv에 게시한 'R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Large Language Models (MLLMs)#Auto-Thinking#Reinforcement Learning (RL)#Bi-mode Annealing#Bi-mode Policy Optimization (BPO)#General-Purpose AI#Reasoning#Efficiency
댓글 수 로딩 중

[논문리뷰] Persuasion Dynamics in LLMs: Investigating Robustness and Adaptability in Knowledge and Safety with DuET-PD

Roy Ka-Wei Lee이 arXiv에 게시한 'Persuasion Dynamics in LLMs: Investigating Robustness and Adaptability in Knowledge and Safety with DuET-PD' 논문에 대한 자세한 리뷰입니다.
#Review#Persuasion Dynamics#Large Language Models (LLMs)#Robustness#Gullibility#Receptiveness#Direct Preference Optimization (DPO)#Safety Alignment#Multi-turn Dialogue
댓글 수 로딩 중

[논문리뷰] OnGoal: Tracking and Visualizing Conversational Goals in Multi-Turn Dialogue with Large Language Models

Alex Endert이 arXiv에 게시한 'OnGoal: Tracking and Visualizing Conversational Goals in Multi-Turn Dialogue with Large Language Models' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models (LLMs)#Human-Computer Interaction (HCI)#Conversational AI#Goal Tracking#Visualization#Multi-Turn Dialogue#User Interface Design#Sensemaking
댓글 수 로딩 중

[논문리뷰] MotionFlux: Efficient Text-Guided Motion Generation through Rectified Flow Matching and Preference Alignment

An-An Liu이 arXiv에 게시한 'MotionFlux: Efficient Text-Guided Motion Generation through Rectified Flow Matching and Preference Alignment' 논문에 대한 자세한 리뷰입니다.
#Review#Text-Guided Motion Generation#Rectified Flow Matching#Preference Alignment#Human Motion Synthesis#Real-time AI#Transformer Architecture#Self-supervised Learning
댓글 수 로딩 중

[논문리뷰] MIDAS: Multimodal Interactive Digital-human Synthesis via Real-time Autoregressive Video Generation

Yan Zhou이 arXiv에 게시한 'MIDAS: Multimodal Interactive Digital-human Synthesis via Real-time Autoregressive Video Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Generation#Digital Human Synthesis#Real-time Video Generation#Autoregressive LLM#Diffusion Models#Deep Compression Autoencoder#Exposure Bias Mitigation#Streaming Inference
댓글 수 로딩 중

[논문리뷰] CODA: Coordinating the Cerebrum and Cerebellum for a Dual-Brain Computer Use Agent with Decoupled Reinforcement Learning

Jianze Liang이 arXiv에 게시한 'CODA: Coordinating the Cerebrum and Cerebellum for a Dual-Brain Computer Use Agent with Decoupled Reinforcement Learning' 논문에 대한 자세한 리뷰입니다.
#Review#GUI Agents#Reinforcement Learning#Planner-Executor Architecture#Decoupled Training#Large Vision-Language Models#Specialization#Generalization#Computer Use Agent
댓글 수 로딩 중

[논문리뷰] TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling

Zhoufutu Wen이 arXiv에 게시한 'TreePO: Bridging the Gap of Policy Optimization and Efficacy and Inference Efficiency with Heuristic Tree-based Modeling' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Policy Optimization#Large Language Models#Inference Efficiency#Tree Search#Segment-level Decoding#Advantage Estimation#Reasoning
댓글 수 로딩 중

[논문리뷰] Breaking the Exploration Bottleneck: Rubric-Scaffolded Reinforcement Learning for General LLM Reasoning

Jiale Zhao이 arXiv에 게시한 'Breaking the Exploration Bottleneck: Rubric-Scaffolded Reinforcement Learning for General LLM Reasoning' 논문에 대한 자세한 리뷰입니다.
#Review#Reinforcement Learning#Large Language Models#Exploration Bottleneck#Instructional Scaffolding#Rubric-based Rewards#General Reasoning#RL with Verifiable Rewards#Policy Optimization
댓글 수 로딩 중

[논문리뷰] On-Policy RL Meets Off-Policy Experts: Harmonizing Supervised Fine-Tuning and Reinforcement Learning via Dynamic Weighting

Guoyin Wang이 arXiv에 게시한 'On-Policy RL Meets Off-Policy Experts: Harmonizing Supervised Fine-Tuning and Reinforcement Learning via Dynamic Weighting' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models#Reinforcement Learning#Supervised Fine-Tuning#On-Policy RL#Off-Policy Experts#Dynamic Weighting#LLM Alignment#Reasoning
댓글 수 로딩 중

[논문리뷰] CorrSteer: Steering Improves Task Performance and Safety in LLMs through Correlation-based Sparse Autoencoder Feature Selection

Adriano Koshiyama이 arXiv에 게시한 'CorrSteer: Steering Improves Task Performance and Safety in LLMs through Correlation-based Sparse Autoencoder Feature Selection' 논문에 대한 자세한 리뷰입니다.
#Review#Sparse Autoencoders#LLM Steering#Feature Selection#Correlation Analysis#AI Safety#Bias Mitigation#Mechanistic Interpretability
댓글 수 로딩 중

[논문리뷰] Copyright Protection for Large Language Models: A Survey of Methods, Challenges, and Trends

Xixiang Zhao이 arXiv에 게시한 'Copyright Protection for Large Language Models: A Survey of Methods, Challenges, and Trends' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Copyright Protection#Model Fingerprinting#Text Watermarking#Invasive Fingerprinting#Intrinsic Fingerprinting#Intellectual Property#Digital Rights Management#Backdoor Watermarking
댓글 수 로딩 중

[논문리뷰] From Black Box to Transparency: Enhancing Automated Interpreting Assessment with Explainable AI in College Classrooms

Ziyin Zhang이 arXiv에 게시한 'From Black Box to Transparency: Enhancing Automated Interpreting Assessment with Explainable AI in College Classrooms' 논문에 대한 자세한 리뷰입니다.
#Review#Automated Interpreting Assessment#Explainable AI#Data Augmentation#Variational Autoencoder#SHAP#Interpreting Quality#Natural Language Processing
댓글 수 로딩 중

[논문리뷰] When Explainability Meets Privacy: An Investigation at the Intersection of Post-hoc Explainability and Differential Privacy in the Context of Natural Language Processing

Gjergji Kasneci이 arXiv에 게시한 'When Explainability Meets Privacy: An Investigation at the Intersection of Post-hoc Explainability and Differential Privacy in the Context of Natural Language Processing' 논문에 대한 자세한 리뷰입니다.
#Review#Natural Language Processing (NLP)#Explainable AI (XAI)#Post-hoc Explainability#Differential Privacy (DP)#Privacy-Utility Trade-off#Model Faithfulness#Text Privatization
댓글 수 로딩 중

[논문리뷰] Can LLM-Generated Textual Explanations Enhance Model Classification Performance? An Empirical Study

Gjergji Kasneci이 arXiv에 게시한 'Can LLM-Generated Textual Explanations Enhance Model Classification Performance? An Empirical Study' 논문에 대한 자세한 리뷰입니다.
#Review#Explainable NLP#Natural Language Explanations#Large Language Models#Pre-trained Language Models#Natural Language Inference#Model Performance Enhancement#Text Generation
댓글 수 로딩 중

[논문리뷰] WGAST: Weakly-Supervised Generative Network for Daily 10 m Land Surface Temperature Estimation via Spatio-Temporal Fusion

Rachid Nedjai이 arXiv에 게시한 'WGAST: Weakly-Supervised Generative Network for Daily 10 m Land Surface Temperature Estimation via Spatio-Temporal Fusion' 논문에 대한 자세한 리뷰입니다.
#Review#Spatio-Temporal Fusion#Land Surface Temperature#Generative Adversarial Network#Weakly-Supervised Learning#Remote Sensing#Deep Learning
댓글 수 로딩 중

[논문리뷰] UNCAGE: Contrastive Attention Guidance for Masked Generative Transformers in Text-to-Image Generation

Kevin Galim이 arXiv에 게시한 'UNCAGE: Contrastive Attention Guidance for Masked Generative Transformers in Text-to-Image Generation' 논문에 대한 자세한 리뷰입니다.
#Review#Text-to-Image Generation#Masked Generative Transformers#Compositional Generation#Attention Guidance#Unmasking Strategy#Contrastive Learning#Training-Free#Attribute Binding
댓글 수 로딩 중

[논문리뷰] Bridging Theory and Practice in Quantum Game Theory: Optimized Implementation of the Battle of the Sexes with Error Mitigation on NISQ Hardware

Jhon Alejandro Andrade이 arXiv에 게시한 'Bridging Theory and Practice in Quantum Game Theory: Optimized Implementation of the Battle of the Sexes with Error Mitigation on NISQ Hardware' 논문에 대한 자세한 리뷰입니다.
#Review#Quantum Game Theory#NISQ Hardware#Error Mitigation#Battle of the Sexes#Qiskit#Quantum Computing#Strategic Coordination#Payoff Maximization
댓글 수 로딩 중

[논문리뷰] VisR-Bench: An Empirical Study on Visual Retrieval-Augmented Generation for Multilingual Long Document Understanding

Tong Yu이 arXiv에 게시한 'VisR-Bench: An Empirical Study on Visual Retrieval-Augmented Generation for Multilingual Long Document Understanding' 논문에 대한 자세한 리뷰입니다.
#Review#Multimodal Retrieval#Retrieval-Augmented Generation#Long Document Understanding#Multilingual NLP#Visual QA#Benchmark#MLLMs#Table Understanding
댓글 수 로딩 중

[논문리뷰] A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems

Xinhao Yi이 arXiv에 게시한 'A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems' 논문에 대한 자세한 리뷰입니다.
#Review#Self-Evolving AI Agents#Lifelong Learning#Foundation Models#Multi-Agent Systems#Agent Optimization#Prompt Engineering#Tool Use#AI Safety#Survey
댓글 수 로딩 중

[논문리뷰] Visual Document Understanding and Question Answering: A Multi-Agent Collaboration Framework with Test-Time Scaling

Ruolin Shen이 arXiv에 게시한 'Visual Document Understanding and Question Answering: A Multi-Agent Collaboration Framework with Test-Time Scaling' 논문에 대한 자세한 리뷰입니다.
#Review#Visual Document Understanding#Visual Question Answering#Multi-Agent System#Test-Time Scaling#Self-Correction#Mixed Reward Modeling#Large Language Models
댓글 수 로딩 중

[논문리뷰] InfiAlign: A Scalable and Sample-Efficient Framework for Aligning LLMs to Enhance Reasoning Capabilities

Zhijie Sang이 arXiv에 게시한 'InfiAlign: A Scalable and Sample-Efficient Framework for Aligning LLMs to Enhance Reasoning Capabilities' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Alignment#Reasoning#Data Curation#Supervised Fine-tuning (SFT)#Direct Preference Optimization (DPO)#Sample Efficiency#Scalability#Multi-dimensional Filtering
댓글 수 로딩 중

[논문리뷰] Can Large Multimodal Models Actively Recognize Faulty Inputs? A Systematic Evaluation Framework of Their Input Scrutiny Ability

Yuan Wu이 arXiv에 게시한 'Can Large Multimodal Models Actively Recognize Faulty Inputs? A Systematic Evaluation Framework of Their Input Scrutiny Ability' 논문에 대한 자세한 리뷰입니다.
#Review#Large Multimodal Models#Input Scrutiny#Error Detection#Faulty Inputs#Evaluation Framework#Modality Preference#Cross-Modal Inconsistency
댓글 수 로딩 중

[논문리뷰] RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization

Kechi Zhang이 arXiv에 게시한 'RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization' 논문에 대한 자세한 리뷰입니다.
#Review#Large Language Models#Reinforcement Learning#Capability Collapse#Hybrid Policy Optimization#Multiple Importance Sampling#Exploration#Math Reasoning#Out-of-Distribution
댓글 수 로딩 중

[논문리뷰] TRACEALIGN -- Tracing the Drift: Attributing Alignment Failures to Training-Time Belief Sources in LLMs

Aman Chadha이 arXiv에 게시한 'TRACEALIGN -- Tracing the Drift: Attributing Alignment Failures to Training-Time Belief Sources in LLMs' 논문에 대한 자세한 리뷰입니다.
#Review#LLM Alignment#Alignment Drift#Training Data Provenance#Belief Conflict Index (BCI)#Suffix Array#Safety Interventions#Reinforcement Learning from Human Feedback#Explainable AI
댓글 수 로딩 중

[논문리뷰] AlignGuard-LoRA: Alignment-Preserving Fine-Tuning via Fisher-Guided Decomposition and Riemannian-Geodesic Collision Regularization

Aman Chadha이 arXiv에 게시한 'AlignGuard-LoRA: Alignment-Preserving Fine-Tuning via Fisher-Guided Decomposition and Riemannian-Geodesic Collision Regularization' 논문에 대한 자세한 리뷰입니다.
#Review#Alignment Preservation#Fine-Tuning#LoRA#Fisher Information Matrix#Catastrophic Forgetting#LLM Safety#Riemannian Geometry#Parameter-Efficient Learning
댓글 수 로딩 중

[논문리뷰] Beyond Linear Bottlenecks: Spline-Based Knowledge Distillation for Culturally Diverse Art Style Classification

Abdelmalik Taleb-Ahmed이 arXiv에 게시한 'Beyond Linear Bottlenecks: Spline-Based Knowledge Distillation for Culturally Diverse Art Style Classification' 논문에 대한 자세한 리뷰입니다.
#Review#Kolmogorov-Arnold Networks#Knowledge Distillation#Art Style Classification#Self-Supervised Learning#Spline-Based Activation#Dual-Teacher#Gram Matrix
댓글 수 로딩 중