AI研究论文每日精选

Analyze Feature Flow to Enhance Interpretation and Steering in Language Models

Gold-medalist Performance in Solving Olympiad Geometry with AlphaGeometry2

ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features

Great Models Think Alike and this Undermines AI Oversight

Ola: Pushing the Frontiers of Omni-Modal Language Model with Progressive Modality Alignment

DynVFX: Augmenting Real Videos with Dynamic Content

Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis

BOLT: Bootstrap Long Chain-of-Thought in Language Models without Distillation

UltraIF: Advancing Instruction Following from the Wild

Weak-to-Strong Diffusion with Reflection

MAGA: MAssive Genre-Audience Reformulation to Pretraining Corpus Expansion

ScoreFlow: Mastering LLM Agent Workflows via Score-based Preference Optimization

MotionCanvas: Cinematic Shot Design with Controllable Image-to-Video Generation

MotionLab: Unified Human Motion Generation and Editing via the Motion-Condition-Motion Paradigm

Beyond Prompt Content: Enhancing LLM Performance via Content-Format Integrated Prompt Optimization

PILAF: Optimal Human Preference Sampling for Reward Modeling

Towards Physical Understanding in Video Generation: A 3D Point Regularization Approach

ChartCitor: Multi-Agent Framework for Fine-Grained Chart Visual Attribution

Learning Real-World Action-Video Dynamics with Heterogeneous Masked Autoregression

PlotGen: Multi-Agent LLM-based Scientific Data Visualization via Multimodal Feedback

Enhancing Code Generation for Low-Resource Languages: No Silver Bullet

Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions