Review

[논문리뷰] AgentFrontier: Expanding the Capability Frontier of LLM Agents with ZPD-Guided Data Synthesis

arXiv에 게시된 'AgentFrontier: Expanding the Capability Frontier of LLM Agents with ZPD-Guided Data Synthesis' 논문에 대한 자세한 리뷰입니다.

#Review #LLM Agents #Data Synthesis #Zone of Proximal Development (ZPD)#Complex Reasoning #Tool Use #Automated Benchmarking #Agentic AI #Rejection Sampling Fine-Tuning

2025년 10월 29일

[논문리뷰] AgentFold: Long-Horizon Web Agents with Proactive Context Management

arXiv에 게시된 'AgentFold: Long-Horizon Web Agents with Proactive Context Management' 논문에 대한 자세한 리뷰입니다.

#Review #Web Agents #Context Management #Long-Horizon Tasks #LLM #Deep Consolidation #Granular Condensation #ReAct Paradigm

2025년 10월 29일

[논문리뷰] ATLAS: Adaptive Transfer Scaling Laws for Multilingual Pretraining, Finetuning, and Decoding the Curse of Multilinguality

arXiv에 게시된 'ATLAS: Adaptive Transfer Scaling Laws for Multilingual Pretraining, Finetuning, and Decoding the Curse of Multilinguality' 논문에 대한 자세한 리뷰입니다.

#Review #Multilingual LLMs #Scaling Laws #Transfer Learning #Curse of Multilinguality #Pretraining #Finetuning #Language Models #Adaptive Scaling

2025년 10월 29일

[논문리뷰] VoMP: Predicting Volumetric Mechanical Property Fields

arXiv에 게시된 'VoMP: Predicting Volumetric Mechanical Property Fields' 논문에 대한 자세한 리뷰입니다.

#Review #Volumetric Properties #Mechanical Simulation #Material Prediction #3D Representation #Physics-based AI #Variational Autoencoder #Geometry Transformer #Gaussian Splats

2025년 10월 28일

[논문리뷰] VITA-E: Natural Embodied Interaction with Concurrent Seeing, Hearing, Speaking, and Acting

Haihan Gao이 arXiv에 게시한 'VITA-E: Natural Embodied Interaction with Concurrent Seeing, Hearing, Speaking, and Acting' 논문에 대한 자세한 리뷰입니다.

#Review #Embodied AI #Human-Robot Interaction #Vision-Language Models #Concurrency #Interruption #Robotics Control #Dual-Model Architecture #Special Tokens

2025년 10월 28일

[논문리뷰] Track, Inpaint, Resplat: Subject-driven 3D and 4D Generation with Progressive Texture Infilling

Igor Gilitschenski이 arXiv에 게시한 'Track, Inpaint, Resplat: Subject-driven 3D and 4D Generation with Progressive Texture Infilling' 논문에 대한 자세한 리뷰입니다.

#Review #Subject-driven 3D/4D Generation #Texture Infilling #Video Tracking #Image Inpainting #Multi-view Consistency #Identity Preservation #Generative Models #3D Gaussians

2025년 10월 28일

[논문리뷰] The Best of N Worlds: Aligning Reinforcement Learning with Best-of-N Sampling via max@k Optimisation

arXiv에 게시된 'The Best of N Worlds: Aligning Reinforcement Learning with Best-of-N Sampling via max@k Optimisation' 논문에 대한 자세한 리뷰입니다.

#Review #Reinforcement Learning #Large Language Models #Best-of-N Sampling #Max@k Optimization #Policy Gradients #Off-policy Learning #Code Generation

2025년 10월 28일

[논문리뷰] RobotArena infty: Scalable Robot Benchmarking via Real-to-Sim Translation

Kuan-Hsun Tu이 arXiv에 게시한 'RobotArena infty: Scalable Robot Benchmarking via Real-to-Sim Translation' 논문에 대한 자세한 리뷰입니다.

#Review #Robot Benchmarking #Real-to-Sim Translation #Vision-Language Models (VLMs)#Human Preference Learning #Domain Randomization #Robot Manipulation #Simulation Environments #Policy Evaluation

2025년 10월 28일

[논문리뷰] ReCode: Unify Plan and Action for Universal Granularity Control

Yifan Wu이 arXiv에 게시한 'ReCode: Unify Plan and Action for Universal Granularity Control' 논문에 대한 자세한 리뷰입니다.

#Review #LLM Agents #Decision Granularity Control #Recursive Code Generation #Hierarchical Planning #Action Unification #Program Synthesis #Data Efficiency

2025년 10월 28일

[논문리뷰] PixelRefer: A Unified Framework for Spatio-Temporal Object Referring with Arbitrary Granularity

Kehan Li이 arXiv에 게시한 'PixelRefer: A Unified Framework for Spatio-Temporal Object Referring with Arbitrary Granularity' 논문에 대한 자세한 리뷰입니다.

#Review #MLLM #Region-level Understanding #Object-centric Reasoning #Spatio-temporal Referring #Video Understanding #Scale-Adaptive Tokenizer #Efficiency #Instruction Tuning

2025년 10월 28일

[논문리뷰] Omni-Reward: Towards Generalist Omni-Modal Reward Modeling with Free-Form Preferences

arXiv에 게시된 'Omni-Reward: Towards Generalist Omni-Modal Reward Modeling with Free-Form Preferences' 논문에 대한 자세한 리뷰입니다.

#Review #Reward Modeling #Multimodal AI #Human Preferences #RLHF #Generalist AI #Benchmark #Dataset #Free-Form Preferences

2025년 10월 28일

[논문리뷰] Mitigating Attention Sinks and Massive Activations in Audio-Visual Speech Recognition with LLMS

arXiv에 게시된 'Mitigating Attention Sinks and Massive Activations in Audio-Visual Speech Recognition with LLMS' 논문에 대한 자세한 리뷰입니다.

#Review #Audio-Visual Speech Recognition #Large Language Models #Attention Sinks #Massive Activations #Decorrelation Loss #Fine-tuning #Multimodal AI

2025년 10월 28일

[논문리뷰] Memory-based Language Models: An Efficient, Explainable, and Eco-friendly Approach to Large Language Modeling

arXiv에 게시된 'Memory-based Language Models: An Efficient, Explainable, and Eco-friendly Approach to Large Language Modeling' 논문에 대한 자세한 리뷰입니다.

#Review #Memory-based Language Model #k-Nearest Neighbor #Eco-friendly AI #Explainable AI #Next-token Prediction #Prefix Trie #Low-latency Inference #CPU-based AI

2025년 10월 28일

[논문리뷰] MARS-M: When Variance Reduction Meets Matrices

arXiv에 게시된 'MARS-M: When Variance Reduction Meets Matrices' 논문에 대한 자세한 리뷰입니다.

#Review #Variance Reduction #Matrix-based Optimizer #LLM Training #Deep Learning Optimization #Moonlight #MARS-M #Stochastic Gradient Descent

2025년 10월 28일

[논문리뷰] Lookahead Anchoring: Preserving Character Identity in Audio-Driven Human Animation

Honglie Chen이 arXiv에 게시한 'Lookahead Anchoring: Preserving Character Identity in Audio-Driven Human Animation' 논문에 대한 자세한 리뷰입니다.

#Review #Audio-driven Animation #Identity Preservation #Diffusion Transformers #Long-form Video Generation #Temporal Autoregression #Keyframe Anchoring #Self-keyframing

2025년 10월 28일

[논문리뷰] LongCat-Video Technical Report

Hongyu Li이 arXiv에 게시한 'LongCat-Video Technical Report' 논문에 대한 자세한 리뷰입니다.

#Review #Video Generation #Diffusion Transformer #RLHF #Sparse Attention #Long Video Generation #Coarse-to-Fine Generation #Multi-task Learning #World Models

2025년 10월 28일

[논문리뷰] LimRank: Less is More for Reasoning-Intensive Information Reranking

Arman Cohan이 arXiv에 게시한 'LimRank: Less is More for Reasoning-Intensive Information Reranking' 논문에 대한 자세한 리뷰입니다.

#Review #Information Reranking #Large Language Models #Data Synthesis #Reasoning-Intensive Retrieval #Low-Resource Learning #Data Efficiency #Instruction Following

2025년 10월 28일

[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

2025년 10월 28일

[논문리뷰] Language Server CLI Empowers Language Agents with Process Rewards

Lanser Contributors이 arXiv에 게시한 'Language Server CLI Empowers Language Agents with Process Rewards' 논문에 대한 자세한 리뷰입니다.

#Review #Language Agents #Language Server Protocol (LSP)#CLI #Process Rewards #Code Refactoring #Static Analysis #Reinforcement Learning #Deterministic Execution

2025년 10월 28일

[논문리뷰] Knocking-Heads Attention

Jianguo Li이 arXiv에 게시한 'Knocking-Heads Attention' 논문에 대한 자세한 리뷰입니다.

#Review #Multi-Head Attention #Transformer #Large Language Models #Inter-Head Communication #Parameter Sharing #Training Stability #Diagonal Initialization

2025년 10월 28일