AI研究論文每日精選

Seed1.5-VL Technical Report

MiMo: Unlocking the Reasoning Potential of Language Model -- From Pretraining to Posttraining

Step1X-3D: Towards High-Fidelity and Controllable Generation of Textured 3D Assets

Learning from Peers in Reasoning Models

Unified Continuous Generative Models

REFINE-AF: A Task-Agnostic Framework to Align Language Models via Self-Generated Instructions using Reinforcement Learning from Automated Feedback

DanceGRPO: Unleashing GRPO on Visual Generation

AttentionInfluence: Adopting Attention Head Influence for Weak-to-Strong Pretraining Data Selection

Skywork-VL Reward: An Effective Reward Model for Multimodal Understanding and Reasoning

WebGen-Bench: Evaluating LLMs on Generating Interactive and Functional Websites from Scratch

Learning Dynamics in Continual Pre-Training for Large Language Models

Reinforced Internal-External Knowledge Synergistic Reasoning for Efficient Adaptive Search Agent

MonetGPT: Solving Puzzles Enhances MLLMs' Image Retouching Skills

H^{3}DP: Triply-Hierarchical Diffusion Policy for Visuomotor Learning

UMoE: Unifying Attention and FFN with Shared Experts

Position: AI Competitions Provide the Gold Standard for Empirical Rigor in GenAI Evaluation

Continuous Visual Autoregressive Generation via Score Maximization

Overflow Prevention Enhances Long-Context Recurrent LLMs

Document Attribution: Examining Citation Relationships using Large Language Models

DynamicRAG: Leveraging Outputs of Large Language Model as Feedback for Dynamic Reranking in Retrieval-Augmented Generation

Physics-Assisted and Topology-Informed Deep Learning for Weather Prediction

LLAMAPIE: Proactive In-Ear Conversation Assistants

Multi-Objective-Guided Discrete Flow Matching for Controllable Biological Sequence Design