arxiv 论文锐评

Attention Is All You Need

Deep Residual Learning for Image Recognition

MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models

Denoising Diffusion Probabilistic Models

Absolute Zero: Reinforced Self-play Reasoning with Zero Data

Mean Flows for One-step Generative Modeling

Who Wrote This? The Key to Zero-Shot LLM-Generated Text Detection Is GECScore

OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments

DetectRL: Benchmarking LLM-Generated Text Detection in Real-World Scenarios

WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation