Tong Yang

The paper introduces Harness-Bench, a diagnostic benchmark that measures how different system 'harnesses' affect LLM agent performance in realistic workflows, showing that agent capability must be reported at the model-harness configuration level.

ESPO is a novel reinforcement learning algorithm that detects trajectory failure in large language models and terminates rollouts early, significantly improving performance on mathematical reasoning benchmarks while reducing computational cost.

ConMoE proposes a train-free method for compressing Mixture-of-Experts (MoE) models by consolidating the large expert pool into a smaller set of reusable prototypes and deterministically remapping all original expert calls to these prototypes.

This paper demonstrates that transformer-based policies can provably learn complex tree search mechanisms, such as depth-first search, purely through reinforcement learning in a stochastic environment.

This paper synthesizes over 150 scattered studies and reports to provide the first comprehensive primer on post-training reasoning data, organizing the field around data objects, utility, construction, and scalability.