AI研究論文每日精選

Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference

AniDoc: Animation Creation Made Easier

TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks

No More Adam: Learning Rate Scaling at Initialization is All You Need

Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces

GUI Agents: A Survey

Mix-LN: Unleashing the Power of Deeper Layers by Combining Pre-LN and Post-LN

LLaVA-UHD v2: an MLLM Integrating High-Resolution Feature Pyramid via Hierarchical Window Transformer

FashionComposer: Compositional Fashion Image Generation

Autoregressive Video Generation without Vector Quantization

FastVLM: Efficient Vision Encoding for Vision Language Models

Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation

AnySat: An Earth Observation Model for Any Resolutions, Scales, and Modalities

Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning

Learning from Massive Human Videos for Universal Humanoid Pose Control

RAG-RewardBench: Benchmarking Reward Models in Retrieval Augmented Generation for Preference Alignment

ChatDiT: A Training-Free Baseline for Task-Agnostic Free-Form Chatting with Diffusion Transformers

VidTok: A Versatile and Open-Source Video Tokenizer

Alignment faking in large language models

CAD-Recode: Reverse Engineering CAD Code from Point Clouds

Predicting the Original Appearance of Damaged Historical Documents

AntiLeak-Bench: Preventing Data Contamination by Automatically Constructing Benchmarks with Updated Real-World Knowledge