Expert-written natural language skills can improve tool-using agents, yet agent-authored skills perform 8-11 points worse than using no skill. This gap suggests that following procedural guidance and improving it from execution evidence are distinct capabilities. Inference time loops can repair skills but do not improv...
Kang Peng, Zhi-Wei Zhang, Yichen Zhang et al.· 1 citation
This work introduces ROSS (Relearning from Self-Generated Rollouts through Selective Supervision), which preserves the full historical trajectory as context while applying loss only to selected model-generated continuations.
Zhi-Wei Zhang, Hua-Yu Deng, Fei Zhao et al.· 0 citations
Experiments show that UEWM provides reliable and transferable reward signals across domains, and that WMG-RL enables a compact 1.7B student policy to match or surpass much larger LLMs on downstream recommendation tasks.
Teacher-Gated On-Policy Distillation is introduced, built on the principle that teacher reliability should be verified at the prompt level before dense supervision is admitted, and outperforms Vanilla OPD in all six single-domain settings and achieves higher seven-benchmark averages at both scales under multi-domain tr...
Zhi-Wei Zhang, Zechen Sun, Fei Zhao et al.· 6 citations
RefVideo-6M, a large-scale reference-guided editing dataset containing 5 million video editing samples and 1 million image editing samples, is introduced and enables the training of powerful editing models with improved visual quality, controllability, and reference consistency.
Bo-Jia Zi, Xiaoyan Yang, Yu Zhou et al.· 1 citation
A novel method that dynamically leverages target-related statements for conversational stance detection by employing a stepwise, entropy-guided backtracking mechanism to selectively activate memory from historical conversations and dynamically constructs a target-aware graph to model the stance relations among utteranc...
Yi-Fan Xiang, Bin Liang, Yu-Qi Huang et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
