A systematic empirical analysis demonstrates that standard metrics including COMET, XCOMET, and BERTScore exhibit near-zero or negative correlation with human cultural judgments, and even display a severity inversion in which scores increase as translation quality deteriorates, and proposes CuRIL, a reinforcement learn...
Yi-Wen Qiu, Linjuan Wu, Ding-Ming Li et al.· 0 citations
GSM8K-V is introduced, a benchmark transforming GSM8K into multi-image sequences with semantic equivalence preserved and error analysis shows that the primary bottleneck lies in Implicit Visual Inference Error (IVIE), where models fail to recover visual semantics that are implied rather than explicitly stated.
Fan Yuan, Yuchen Yan, Yi-Fan Jiang et al.· arXiv.org· 8 citations
This work presents EasySteer, a unified framework for high-performance, extensible LLM steering built on vLLM, and demonstrates its effectiveness in overthinking mitigation, hallucination reduction, and other key applications.
Haolei Xu, Xinyu Mei, Yuchen Yan et al.· arXiv.org· 14 citations
Test-Time Policy Optimization is proposed, an asymmetric objective that distills agreeing rollouts via OPSD and penalizes disagreeing rollouts with Grouped RL and Token-level selection further refines both branches: distillation down-weights already-converged positions, while RL penalizes only confident errors.
Ao-Han Wang, Zhengxi Lu, Jian-Ze Wang et al.· 2 citations
This work introduces PaperGym, a unified framework that turns each research paper into a complete training environment, and releases the pipeline, the 20,000-instance corpus PaperGym-20k, and the benchmarks PaperGym-Innov and PaperGym-Design.
Yu-Han Wang, Zhengxi Lu, Yuchen Yan et al.· 2 citations
This work proposes AgentOPSD, a critic-free, recursive method for turn-level credit assignment in agentic reinforcement learning that aggregates token-level teacher-student log-probability gaps into turn-level evidence and recursively updates a Bayesian belief state in log-odds space.
Zi-Han Wang, Zhengxi Lu, Zhiyuan Yao et al.· 7 citations· ⚡1
EMPO achieves substantial gains over the base model and achieves competitive performance against strong baselines such as UI-TARS-7B and GPT-4o, demonstrating better generalization than prior single-turn RL approaches.
Zhengxi Lu, Jiabo Ye, Fei Tang et al.· Annual Meeting of the Associ...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
