Researchers introduced Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization (SA-MRPO), a method for multi-objective reinforcement learning that normalizes each…
Researchers built AutoResearchEval, a benchmark of 100 frontier research tasks spanning seven scientific domains and the full research lifecycle, and…
Researchers introduced MMDiff, a framework for discovering and controlling interpretable features in multimodal large language models by comparing sparse autoencoders…
Researchers identified and fixed five deployment bugs that prevented the Nanbeige4.2-3B language model from running correctly on Apple Silicon, including…
Researchers conducted circuit analyses across 46 tasks spanning four cognitive domains to test whether large language models develop specialized internal…
Researchers from the University of Chinese Academy of Sciences and Meituan introduced an evaluation framework that assesses AI research agents…
A new paper introduces Second Thought, a training-free inference framework that has LLM agents spawn auxiliary reasoning branches during the…
A study tested whether committees of LLM agents used for clinical decision support can be manipulated through benchmark shortcuts that…
Researchers proposed S²VOPD, a self-supervised on-policy distillation method that creates teacher-student asymmetry by degrading the student’s input rather than augmenting…
SimpleOPD is a new on-policy distillation method for transferring long-context reasoning ability from large teacher models to smaller, shorter-context student…
A paper introduces LOPD, a method that makes an agent’s own privileged context end-to-end learnable from experience rather than relying…
Researchers from Peking University and BUPT identified a failure mode in sequential reinforcement learning with verifiable rewards, where optimizing for…
