6 papers
DeltaTTG: Training Multi-Turn Agents with a Token-to-Go Progress Meter
Xuyan Ye, Xiyan Fu, Shengda Fan +3
Reinforcement learning with verifiable rewards (RLVR) for multi-turn agents relies on sparse terminal rewards that say little about individual steps. Within-group step comparisons…
Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training
Kunlun Zhu, Xuyan Ye, Yibo Li +3
An unsuccessful LLM agent rollout contains more information than its final reward: the observations available to the agent, the actions it chose, and the environment's responses. R…
AgentDebugX: An Open-Source Toolkit for Failure Observability, Attribution, and Recovery in LLM Agents
Kunlun Zhu, Xuyan Ye, Zhiguang Han +9
LLM agent failures are difficult to debug because the step where an error surfaces is often not the one that caused it. Existing observability tools replay execution traces but pro…
The Many Faces of On-Policy Distillation: Pitfalls, Mechanisms, and Fixes
Siqi Zhu, Xuyan Ye, Hongyu Lu +2
On-policy distillation (OPD) and on-policy self-distillation (OPSD) have emerged as promising post-training methods for large language models, offering dense token-level supervisio…
AgentProcessBench: Diagnosing Step-Level Process Quality in Tool-Using Agents
Shengda Fan, Xuyan Ye, Yupeng Huo +9
While Large Language Models (LLMs) have evolved into tool-using agents, they remain brittle in long-horizon interactions. Unlike mathematical reasoning where errors are often recti…
DARC: Decoupled Asymmetric Reasoning Curriculum for LLM Evolution
Shengda Fan, Xuyan Ye, Yankai Lin
Self-play with large language models has emerged as a promising paradigm for achieving self-improving artificial intelligence. However, existing self-play frameworks often suffer f…