Great Minds Think Alike: Contextual Tacit Communication for Decentralized LLM-Agent Cooperation

Yue Pei, Hongming Zhang, Jiarui Guan, Jusheng Zhang, Liang Lin, Haogang Zhu, Ziliang Chen
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:97756-97777, 2026.

Abstract

Large language models (LLMs) are increasingly used as planners for cooperative embodied agents, but multi-agent settings amplify inconsistency under partial observability and make explicit communication costly or even unavailable. Many existing approaches rely on online message passing; when communication is removed, agents often fall back to independent local planning that suffers from miscoordination. We introduce Contextual Tacit Communication, a gradient-free protocol that aligns decentralized decisions with a joint LLM value score without explicit message actions. Our method measures context-conditioned value rectifications via residual banding to pinpoint miscoordination actions and amortizes the resulting coordination signals into a retrieval-augmented Tacit Rule Memory that provides prompt-level cooperation rules at execution time. Experiments on VIKI, C-WAH, and TDW-MAT show that our approach improves cooperation performance over baselines while reducing runtime overhead compared with communication-based methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-pei26d, title = {Great Minds Think Alike: Contextual Tacit Communication for Decentralized {LLM}-Agent Cooperation}, author = {Pei, Yue and Zhang, Hongming and Guan, Jiarui and Zhang, Jusheng and Lin, Liang and Zhu, Haogang and Chen, Ziliang}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {97756--97777}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/pei26d/pei26d.pdf}, url = {https://proceedings.mlr.press/v306/pei26d.html}, abstract = {Large language models (LLMs) are increasingly used as planners for cooperative embodied agents, but multi-agent settings amplify inconsistency under partial observability and make explicit communication costly or even unavailable. Many existing approaches rely on online message passing; when communication is removed, agents often fall back to independent local planning that suffers from miscoordination. We introduce Contextual Tacit Communication, a gradient-free protocol that aligns decentralized decisions with a joint LLM value score without explicit message actions. Our method measures context-conditioned value rectifications via residual banding to pinpoint miscoordination actions and amortizes the resulting coordination signals into a retrieval-augmented Tacit Rule Memory that provides prompt-level cooperation rules at execution time. Experiments on VIKI, C-WAH, and TDW-MAT show that our approach improves cooperation performance over baselines while reducing runtime overhead compared with communication-based methods.} }
Endnote
%0 Conference Paper %T Great Minds Think Alike: Contextual Tacit Communication for Decentralized LLM-Agent Cooperation %A Yue Pei %A Hongming Zhang %A Jiarui Guan %A Jusheng Zhang %A Liang Lin %A Haogang Zhu %A Ziliang Chen %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-pei26d %I PMLR %P 97756--97777 %U https://proceedings.mlr.press/v306/pei26d.html %V 306 %X Large language models (LLMs) are increasingly used as planners for cooperative embodied agents, but multi-agent settings amplify inconsistency under partial observability and make explicit communication costly or even unavailable. Many existing approaches rely on online message passing; when communication is removed, agents often fall back to independent local planning that suffers from miscoordination. We introduce Contextual Tacit Communication, a gradient-free protocol that aligns decentralized decisions with a joint LLM value score without explicit message actions. Our method measures context-conditioned value rectifications via residual banding to pinpoint miscoordination actions and amortizes the resulting coordination signals into a retrieval-augmented Tacit Rule Memory that provides prompt-level cooperation rules at execution time. Experiments on VIKI, C-WAH, and TDW-MAT show that our approach improves cooperation performance over baselines while reducing runtime overhead compared with communication-based methods.
APA
Pei, Y., Zhang, H., Guan, J., Zhang, J., Lin, L., Zhu, H. & Chen, Z.. (2026). Great Minds Think Alike: Contextual Tacit Communication for Decentralized LLM-Agent Cooperation. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:97756-97777 Available from https://proceedings.mlr.press/v306/pei26d.html.

Related Material