AI研究論文每日精選

Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs

DreamBench++: A Human-Aligned Benchmark for Personalized Image Generation

BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions

Evaluating D-MERIT of Partial-annotation on Information Retrieval

Long Context Transfer from Language to Vision

Video-Infinity: Distributed Long Video Generation

VideoHallucer: Evaluating Intrinsic and Extrinsic Hallucinations in Large Video-Language Models

WARP: On the Benefits of Weight Averaged Rewarded Policies

Scaling Laws for Linear Complexity Language Models

Towards Fast Multilingual LLM Inference: Speculative Decoding and Specialized Drafters

Efficient Continual Pre-training by Mitigating the Stability Gap

Sparser is Faster and Less is More: Efficient Sparse Attention for Long-Range Transformers

Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs

Beyond the Turn-Based Game: Enabling Real-Time Conversations with Duplex Models

Preference Tuning For Toxicity Mitigation Generalizes Across Languages

AutoDetect: Towards a Unified Framework for Automated Weakness Detection in Large Language Models

Confidence Regulation Neurons in Language Models

How Many Parameters Does it Take to Change a Light Bulb? Evaluating Performance in Self-Play of Conversational Games as a Function of Model Characteristics

ClotheDreamer: Text-Guided Garment Generation with 3D Gaussians

Found in the Middle: Calibrating Positional Attention Bias Improves Long Context Utilization

IRASim: Learning Interactive Real-Robot Action Simulators

video-SALMONN: Speech-Enhanced Audio-Visual Large Language Models

Can Few-shot Work in Long-Context? Recycling the Context to Generate Demonstrations

Repulsive Score Distillation for Diverse Sampling of Diffusion Models

OlympicArena Medal Ranks: Who Is the Most Intelligent AI So Far?