Review

[논문리뷰] LightCache: Memory-Efficient, Training-Free Acceleration for Video Generation

Zheng Zhan이 arXiv에 게시한 'LightCache: Memory-Efficient, Training-Free Acceleration for Video Generation' 논문에 대한 자세한 리뷰입니다.

#Review #Video Generation #Diffusion Models #Memory Efficiency #Inference Acceleration #Training-Free #Cache Mechanism #GPU Optimization

2025년 10월 8일

[논문리뷰] Less is More: Recursive Reasoning with Tiny Networks

arXiv에 게시된 'Less is More: Recursive Reasoning with Tiny Networks' 논문에 대한 자세한 리뷰입니다.

#Review #Recursive Reasoning #Tiny Networks #Deep Supervision #Hierarchical Reasoning Model (HRM)#Sudoku-Extreme #ARC-AGI #Generalization #Parameter Efficiency

2025년 10월 8일

[논문리뷰] In-the-Flow Agentic System Optimization for Effective Planning and Tool Use

arXiv에 게시된 'In-the-Flow Agentic System Optimization for Effective Planning and Tool Use' 논문에 대한 자세한 리뷰입니다.

#Review #Agentic Systems #Large Language Models (LLMs)#Tool Use #Reinforcement Learning (RL)#On-policy Optimization #Flow-based Group Refined Policy Optimization (Flow-GRPO)#Multi-turn Reasoning

2025년 10월 8일

[논문리뷰] Human3R: Everyone Everywhere All at Once

Yuliang Xiu이 arXiv에 게시한 'Human3R: Everyone Everywhere All at Once' 논문에 대한 자세한 리뷰입니다.

#Review #4D Human-Scene Reconstruction #Online Reconstruction #Multi-person #SMPL-X #Transformer #Visual Prompt Tuning #Real-time #Foundation Model

2025년 10월 8일

[논문리뷰] HoloScene: Simulation-Ready Interactive 3D Worlds from a Single Video

Katelyn Gao이 arXiv에 게시한 'HoloScene: Simulation-Ready Interactive 3D Worlds from a Single Video' 논문에 대한 자세한 리뷰입니다.

#Review #3D Reconstruction #Digital Twin #Scene Graph #Physical Simulation #Interactive Environments #Single Video Reconstruction #Neural Rendering

2025년 10월 8일

[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

2025년 10월 8일

[논문리뷰] Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval and Synthesis for SLMs

arXiv에 게시된 'Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval and Synthesis for SLMs' 논문에 대한 자세한 리뷰입니다.

#Review #DeepResearch Agents #Tool-integrated Reasoning #Reinforcement Learning #Information Retrieval #Information Synthesis #Multi-agent Self-play #Reward Shaping #LLM

2025년 10월 8일

[논문리뷰] Fast-dLLM v2: Efficient Block-Diffusion LLM

arXiv에 게시된 'Fast-dLLM v2: Efficient Block-Diffusion LLM' 논문에 대한 자세한 리뷰입니다.

#Review #Diffusion LLMs #Inference Acceleration #Parallel Decoding #Autoregressive Models #Caching #Fine-tuning #Block-wise Attention

2025년 10월 8일

[논문리뷰] Equilibrium Matching: Generative Modeling with Implicit Energy-Based Models

arXiv에 게시된 'Equilibrium Matching: Generative Modeling with Implicit Energy-Based Models' 논문에 대한 자세한 리뷰입니다.

#Review #Generative Models #Equilibrium Dynamics #Energy-Based Models (EBMs)#Flow Matching #Diffusion Models #Optimization-Based Sampling #Image Generation

2025년 10월 8일

[논문리뷰] EgoNight: Towards Egocentric Vision Understanding at Night with a Challenging Benchmark

Tianwen Qian이 arXiv에 게시한 'EgoNight: Towards Egocentric Vision Understanding at Night with a Challenging Benchmark' 논문에 대한 자세한 리뷰입니다.

#Review #Egocentric Vision #Nighttime Conditions #Visual Question Answering (VQA)#Day-Night Alignment #Multimodal Large Language Models (MLLMs)#Depth Estimation #Correspondence Retrieval #Benchmark

2025년 10월 8일

[논문리뷰] Drax: Speech Recognition with Discrete Flow Matching

arXiv에 게시된 'Drax: Speech Recognition with Discrete Flow Matching' 논문에 대한 자세한 리뷰입니다.

#Review #Automatic Speech Recognition (ASR)#Discrete Flow Matching (DFM)#Non-Autoregressive (NAR)#Generative Models #Tri-mixture Probability Path #Parallel Decoding #Accuracy-Efficiency Trade-off #Speech Synthesis

2025년 10월 8일

[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

2025년 10월 8일

[논문리뷰] Discrete Diffusion Models with MLLMs for Unified Medical Multimodal Generation

arXiv에 게시된 'Discrete Diffusion Models with MLLMs for Unified Medical Multimodal Generation' 논문에 대한 자세한 리뷰입니다.

#Review #Discrete Diffusion Models #Multimodal Large Language Models (MLLMs)#Medical Image Generation #Medical Report Generation #Multimodal Generation #Medical AI #Cross-modal Alignment

2025년 10월 8일

[논문리뷰] Demystifying deep search: a holistic evaluation with hint-free multi-hop questions and factorised metrics

arXiv에 게시된 'Demystifying deep search: a holistic evaluation with hint-free multi-hop questions and factorised metrics' 논문에 대한 자세한 리뷰입니다.

#Review #Deep Search #Multi-hop Reasoning #Evaluation Benchmark #Retrieval-Augmented Generation #Web Agents #Diagnostic Metrics #Knowledge Utilization #Hint-Free Questions

2025년 10월 8일

[논문리뷰] Deforming Videos to Masks: Flow Matching for Referring Video Segmentation

Chengzu Li이 arXiv에 게시한 'Deforming Videos to Masks: Flow Matching for Referring Video Segmentation' 논문에 대한 자세한 리뷰입니다.

#Review #Referring Video Object Segmentation #Flow Matching #Video Segmentation #Generative Models #Text-to-Video #Continuous Flow #Diffusion Models

2025년 10월 8일

[논문리뷰] DRIFT: Learning from Abundant User Dissatisfaction in Real-World Preference Learning

Zheli Liu이 arXiv에 게시한 'DRIFT: Learning from Abundant User Dissatisfaction in Real-World Preference Learning' 논문에 대한 자세한 리뷰입니다.

#Review #Preference Learning #LLMs #User Feedback #Dissatisfaction Signals #DPO #Iterative Training #RLHF #Exploration

2025년 10월 8일

[논문리뷰] CoDA: Coding LM via Diffusion Adaptation

arXiv에 게시된 'CoDA: Coding LM via Diffusion Adaptation' 논문에 대한 자세한 리뷰입니다.

#Review #Diffusion Language Models #Code Generation #Bidirectional Decoding #Text Infilling #Instruction Tuning #Lightweight Models #TPU Training

2025년 10월 8일

[논문리뷰] CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding

arXiv에 게시된 'CCD: Mitigating Hallucinations in Radiology MLLMs via Clinical Contrastive Decoding' 논문에 대한 자세한 리뷰입니다.

#Review #Multimodal Large Language Models (MLLMs)#Radiology Report Generation (RRG)#Medical Hallucinations #Contrastive Decoding #Training-free Inference #Clinical AI #Visual Question Answering (VQA)

2025년 10월 8일

[논문리뷰] CARE: Cognitive-reasoning Augmented Reinforcement for Emotional Support Conversation

arXiv에 게시된 'CARE: Cognitive-reasoning Augmented Reinforcement for Emotional Support Conversation' 논문에 대한 자세한 리뷰입니다.

#Review #Emotional Support Conversation #Cognitive Reasoning #Reinforcement Learning #Dialogue Generation #Natural Language Processing #Large Language Models #Psychological Support

2025년 10월 8일

[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

2025년 10월 8일