Review

[논문리뷰] Open-o3 Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence

arXiv에 게시된 'Open-o3 Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence' 논문에 대한 자세한 리뷰입니다.

#Review #Video Reasoning #Spatio-Temporal Grounding #Large Multimodal Models #Reinforcement Learning #Chain-of-Thought #Visual Evidence #Dataset Curation

2025년 10월 24일

[논문리뷰] Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall

Sungjin Ahn이 arXiv에 게시한 'Loopholing Discrete Diffusion: Deterministic Bypass of the Sampling Wall' 논문에 대한 자세한 리뷰입니다.

#Review #Discrete Diffusion Models #Sampling Wall #Loopholing #Self-Conditioning #Non-Autoregressive Generation #Text Generation #Language Modeling #Reasoning Tasks

2025년 10월 24일

[논문리뷰] LayerComposer: Interactive Personalized T2I via Spatially-Aware Layered Canvas

arXiv에 게시된 'LayerComposer: Interactive Personalized T2I via Spatially-Aware Layered Canvas' 논문에 대한 자세한 리뷰입니다.

#Review #Text-to-Image Generation #Personalization #Diffusion Models #Interactive Control #Multi-Subject Composition #Layered Canvas #Spatial Control #Image Editing

2025년 10월 24일

[논문리뷰] Investigating Safety Vulnerabilities of Large Audio-Language Models Under Speaker Emotional Variations

arXiv에 게시된 'Investigating Safety Vulnerabilities of Large Audio-Language Models Under Speaker Emotional Variations' 논문에 대한 자세한 리뷰입니다.

#Review #LALM Safety #Speaker Emotion #Safety Alignment #Jailbreaking #Audio-Language Models #Emotional Variation #Unsafe Rate #Non-refusal Rate

2025년 10월 24일

[논문리뷰] ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases

Nicholas Carlini이 arXiv에 게시한 'ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases' 논문에 대한 자세한 리뷰입니다.

#Review #LLM Evaluation #Reward Hacking #Benchmark Reliability #Test Exploitation #Prompt Engineering #LLM Safety #Code Generation

2025년 10월 24일

[논문리뷰] Human-Agent Collaborative Paper-to-Page Crafting for Under $0.1

arXiv에 게시된 'Human-Agent Collaborative Paper-to-Page Crafting for Under $0.1' 논문에 대한 자세한 리뷰입니다.

#Review #Human-Agent Collaboration #Project Page Generation #Multi-Agent System #LLM #VLM #Webpage Automation #PageBench #Scientific Communication #Cost-Effective AI

2025년 10월 24일

[논문리뷰] HoloCine: Holistic Generation of Cinematic Multi-Shot Long Video Narratives

arXiv에 게시된 'HoloCine: Holistic Generation of Cinematic Multi-Shot Long Video Narratives' 논문에 대한 자세한 리뷰입니다.

#Review #Text-to-Video Generation #Multi-Shot Video #Narrative Coherence #Diffusion Models #Self-Attention #Cinematic AI #Video Consistency #Directorial Control

2025년 10월 24일

[논문리뷰] From Masks to Worlds: A Hitchhiker's Guide to World Models

Shufan Li이 arXiv에 게시한 'From Masks to Worlds: A Hitchhiker's Guide to World Models' 논문에 대한 자세한 리뷰입니다.

#Review #World Models #Generative AI #Multimodal Learning #Masked Modeling #Interactive AI #Memory Systems #Autonomous Agents #AI Roadmap

2025년 10월 24일

[논문리뷰] Every Question Has Its Own Value: Reinforcement Learning with Explicit Human Values

arXiv에 게시된 'Every Question Has Its Own Value: Reinforcement Learning with Explicit Human Values' 논문에 대한 자세한 리뷰입니다.

#Review #Reinforcement Learning #LLM Alignment #Human Values #Reward Shaping #Value-Weighted Reward #Termination Policy #RLVR

2025년 10월 24일

[논문리뷰] Emergence of Linear Truth Encodings in Language Models

Alberto Bietti이 arXiv에 게시한 'Emergence of Linear Truth Encodings in Language Models' 논문에 대한 자세한 리뷰입니다.

#Review #Language Models #Truth Encoding #Linear Subspaces #Mechanistic Interpretability #Transformer Models #Learning Dynamics #Truth Co-occurrence Hypothesis #Hallucinations

2025년 10월 24일

[논문리뷰] DyPE: Dynamic Position Extrapolation for Ultra High Resolution Diffusion

arXiv에 게시된 'DyPE: Dynamic Position Extrapolation for Ultra High Resolution Diffusion' 논문에 대한 자세한 리뷰입니다.

#Review #Diffusion Models #Transformer Architecture #Positional Encoding #High-Resolution Image Generation #Extrapolation #Dynamic Adaptation #Training-Free

2025년 10월 24일

[논문리뷰] Diff-XYZ: A Benchmark for Evaluating Diff Understanding

arXiv에 게시된 'Diff-XYZ: A Benchmark for Evaluating Diff Understanding' 논문에 대한 자세한 리뷰입니다.

#Review #Diff Understanding #Code Diff #Benchmark #LLMs #Code Editing #Software Engineering #Unified Diff Format #Search-Replace

2025년 10월 24일

[논문리뷰] Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence

arXiv에 게시된 'Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence' 논문에 대한 자세한 리뷰입니다.

#Review #Video Reasoning #Multimodal Large Language Models (MLLMs)#Reinforcement Learning (RLVR)#Evidence Grounding #Multi-step Reasoning #Frame Retrieval #Dataset Construction #Progressive Learning

2025년 10월 24일

[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)

2025년 10월 24일

[논문리뷰] AlphaFlow: Understanding and Improving MeanFlow Models

arXiv에 게시된 'AlphaFlow: Understanding and Improving MeanFlow Models' 논문에 대한 자세한 리뷰입니다.

#Review #Generative Models #Flow Matching #Consistency Models #MeanFlow #Curriculum Learning #Few-Step Generation #Image Generation

2025년 10월 24일

[논문리뷰] AdaSPEC: Selective Knowledge Distillation for Efficient Speculative Decoders

arXiv에 게시된 'AdaSPEC: Selective Knowledge Distillation for Efficient Speculative Decoders' 논문에 대한 자세한 리뷰입니다.

#Review #Speculative Decoding #Knowledge Distillation #LLM Inference #Model Acceleration #Token Filtering #Draft Model #Acceptance Rate

2025년 10월 24일

[논문리뷰] ARGenSeg: Image Segmentation with Autoregressive Image Generation Model

arXiv에 게시된 'ARGenSeg: Image Segmentation with Autoregressive Image Generation Model' 논문에 대한 자세한 리뷰입니다.

#Review #Image Segmentation #Autoregressive Generation #Multimodal Large Language Models (MLLMs)#Visual Understanding #VQ-VAE #Multi-scale Prediction #Referring Expression Segmentation #Image Generation

2025년 10월 24일

[논문리뷰] olmOCR 2: Unit Test Rewards for Document OCR

arXiv에 게시된 'olmOCR 2: Unit Test Rewards for Document OCR' 논문에 대한 자세한 리뷰입니다.

#Review #Document OCR #Vision Language Model #Reinforcement Learning #Unit Tests #Synthetic Data Generation #RLVR #Document Parsing #State-of-the-Art OCR

2025년 10월 23일

[논문리뷰] VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos

Xinyuan Wang이 arXiv에 게시한 'VideoAgentTrek: Computer Use Pretraining from Unlabeled Videos' 논문에 대한 자세한 리뷰입니다.

#Review #GUI Agents #Video Pretraining #Inverse Dynamics #Action Recognition #Computer Use Automation #Data Synthesis #Multimodal Learning

2025년 10월 23일

[논문리뷰] Unified Reinforcement and Imitation Learning for Vision-Language Models

arXiv에 게시된 'Unified Reinforcement and Imitation Learning for Vision-Language Models' 논문에 대한 자세한 리뷰입니다.

#Review #Vision-Language Models #Reinforcement Learning #Imitation Learning #Model Distillation #Lightweight VLMs #LLM-as-a-Judge #Multimodal Learning

2025년 10월 23일