@article{jiayi2026,
title = {Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents},
author = {Wu, Jiayi and Xie, Ruobing and Huang, Zeqian and Jiang, Lei and Xu, Can and Luo, Kangyang and Lin, Bochen and Gao, Ming and Li, Xiang},
journal = {Findings of the Association for Computational Linguistics: EMNLP 2026},
year = {2026},
month = {Jan},
doi = {10.18653/2026.findings-emnlp.1874},
url = {https://doi.org/10.18653/2026.findings-emnlp.1874},
}
Wu, Jiayi et al. (2026). Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents. Findings of the Association for Computational Linguistics: EMNLP 2026. https://doi.org/10.18653/2026.findings-emnlp.1874
Wu, Jiayi, et al.. "Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents." Findings of the Association for Computational Linguistics: EMNLP 2026, 2026, https://doi.org/10.18653/2026.findings-emnlp.1874.
Wu, Jiayi, et al.. "Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents." Findings of the Association for Computational Linguistics: EMNLP 2026 (2026). https://doi.org/10.18653/2026.findings-emnlp.1874
Wu, Jiayi et al., 2026. Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents. Findings of the Association for Computational Linguistics: EMNLP 2026. Available at: <https://doi.org/10.18653/2026.findings-emnlp.1874>.
TY - JOUR
TI - Negative Advantage Is a Double-Edged Sword: Calibrating advantages in GRPO for Search Agents
AU - Wu, Jiayi
AU - Xie, Ruobing
AU - Huang, Zeqian
AU - Jiang, Lei
AU - Xu, Can
AU - Luo, Kangyang
AU - Lin, Bochen
AU - Gao, Ming
AU - Li, Xiang
PY - 2026
JO - Findings of the Association for Computational Linguistics: EMNLP 2026
DO - 10.18653/2026.findings-emnlp.1874
UR - https://doi.org/10.18653/2026.findings-emnlp.1874
ER -