@article{bipacebisimulationguidedpolicyoptimizati, title = {BiPACE: Bisimulation-Guided Policy Optimization with Action Counterfactual Estimation for LLM Agents}, author = {Hanyang Wang and Weijieying Ren and Yuxiang Zhang and Ding Cao and Zhizhao Zeng and Ke Zeng and Tianxiang Zhao}, year = {2026}, eprint = {2606.25556}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2606.25556}, }