---
title: "Flout at Your Own Risk — paper by VanderHoeven et al. (2026)"
description: "They argue that an utterance is actionable only if it does two things: it shifts the listener's beliefs about the world, and it lowers… 72 references."
authors: ["Hannah VanderHoeven", "Abhijnan Nath", "Nikhil Krishnaswamy"]
year: 2026
venue: "arXiv:2607.11053"
source: "https://arxiv.org/abs/2607.11053"
doi: "10.48550/arxiv.2607.11053"
arxiv: "2607.11053"
licence: "arXiv non-exclusive distribution licence (read only, no redistribution)"
licence_url: "http://arxiv.org/licenses/nonexclusive-distrib/1.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/vanderhoeven-2026-flout-own-risk-llms-struggle/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/vanderhoeven-2026-flout-own-risk-llms-struggle/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/30/) / Flout at Your Own Risk: LLMs Struggle with Pragmatic Cooperativity Under Epistemic Asymmetry

# Flout at Your Own Risk: LLMs Struggle with Pragmatic Cooperativity Under Epistemic Asymmetry

**Authors:** Hannah VanderHoeven, [Abhijnan Nath](https://md.research.elamur.ai/free/en/authors/abhijnan-nath/) and [Nikhil Krishnaswamy](https://md.research.elamur.ai/free/en/authors/nikhil-krishnaswamy/) · **Year:** 2026 · **Venue:** arXiv:2607.11053 · **Licence:** [arXiv non-exclusive distribution licence (read only, no redistribution)](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2607.11053) · [PDF](https://arxiv.org/pdf/2607.11053) · [DOI 10.48550/arxiv.2607.11053](https://doi.org/10.48550/arxiv.2607.11053) · [arXiv 2607.11053](https://arxiv.org/abs/2607.11053) · [OpenAlex](https://openalex.org/W7168268537)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
The paper asks whether LLM agents communicate cooperatively when no participant holds the full picture and the solution only emerges by pooling private knowledge. The authors formalize this setting as collaborative epistemic asymmetry (CEA). They argue that an utterance is actionable only if it does two things: it shifts the listener's beliefs about the world, and it lowers the entropy over the shared goal. They then show that breaking any Gricean maxim undermines one of these two conditions, which turns cooperativity into something measurable through task progress. The testbed is adapted from the DPIP Lego task: three Directors each see one side view of a structure and instruct a single Builder. LLMs are evaluated in both speaker and listener roles, with prompting, move exploration and post-training (SFT, IPO) as interventions. The models show partial pragmatic skill and can sometimes disregard a speaker's lapses. Still, unrecognized floutings, especially of relation, track failed moves. This means claims that LLM collaborators are cooperative deserve skepticism.

## Key points
- The paper poses three research questions: whether LLM listeners behave literally or pragmatically, whether prompting and exploration elicit pragmatic listening, and whether offline alignment makes speakers robust to weak listeners.
- In the DPIP Lego setup, goal structures have a 3 × 3 footprint and are 3 layers tall. Three Directors hold private side views, and one Builder performs PLACE and REMOVE moves. Builder CLARIFY requests are counted as failures.
- A relation flouting occurs when the "owner" of an already completed wall keeps issuing instructions. Such flouting is strongly associated with move failure, which partly explains why games starting from non-empty boards made less progress.
- A manner flouting occurs when one Director gives two instructions in a single turn. A strong Builder, namely GPT-4.1-mini given Pragmatic prompting plus the ability to explore moves, largely tolerates this, with only a slight rise in REMOVE failures, mostly with SFT Directors.
- Aligned open-model Directors improve but stay fragile. Some produce instructions that look cooperative but barely advance the goal, for example placing and then removing the same block.

## Interesting ideas
- Treat each Gricean maxim as a constraint on an utterance and score cooperativity by whether the utterance updates beliefs and reduces goal uncertainty. This replaces subjective judgments of politeness or style with an objective, task-linked signal.
- Operationalize maxims as concrete, checkable events, such as a speaker addressing an already-finished subgoal (relation) or giving two instructions at once (manner). Then correlate each event with outcomes, so abstract pragmatics becomes countable.
- Separate superficially correct actions from goal-advancing ones, such as a place followed by a remove of the same block. Accuracy alone would hide this degenerate kind of "success."
- Count clarification requests as failed moves in the metric. This stops an agent from scoring well by perpetually asking instead of acting.
- Analyze listener robustness to flawed speakers, such as choosing among ambiguous readings or switching to another speaker's utterance. This frames pragmatic listening as selective trust across several sources.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (frames): crate prag could label each maxim violation with the specific condition it breaks (no belief update vs. no goal-entropy drop), making implicature and flouting detection explainable.
- **I2** (world): the world of the text could track which subgoals and states are already complete, so an utterance about a finished state is automatically flagged as non-actionable.
- **I3** (learn): evaluation could include a negative control that detects self-cancelling event sequences, so that trivially correct outputs do not inflate scores during self-play or teacher distillation.
- **I4** (answer): QA readers that abstain or ask back could be scored the same way, keeping the pressure on committing to an answer grounded in the text world.
- **I5** (frames): when several clauses or speakers give conflicting instructions, prag could rank them by actionability against the world state rather than taking them all literally.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works

## Cite

Cite as: Hannah VanderHoeven, Abhijnan Nath and Nikhil Krishnaswamy. 2026. Flout at Your Own Risk: LLMs Struggle with Pragmatic Cooperativity Under Epistemic Asymmetry. arXiv:2607.11053.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{vanderhoeven2026flout,
  title = {{Flout at Your Own Risk: LLMs Struggle with Pragmatic Cooperativity Under Epistemic Asymmetry}},
  author = {Hannah VanderHoeven, Abhijnan Nath and Nikhil Krishnaswamy},
  year = {2026},
  doi = {10.48550/arxiv.2607.11053},
  eprint = {2607.11053},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2607.11053}
}
```

- [← previous Floating or Suggesting Ideas? A Large-Scale Contrastive Analysis of Metaphorical and Literal Verb–Object Constructions](https://md.research.elamur.ai/free/en/papers/piccirilli-2026-floating-or-suggesting-ideas-large/)
- [next → Flow-Based Page Unique Semantic Mapping Architecture for Document Visual Question Answering](https://md.research.elamur.ai/free/en/papers/wang-2026-flow-based-page-unique-semantic/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/30/)

## References

72 references, 18 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Jacob Andreas and Dan Klein. [Reasoning about pragmatics with neural listeners and speakers](https://md.research.elamur.ai/free/en/papers/andreas-2016-reasoning-about-pragmatics-neural-listeners/). In Proceedings of the 2016 conference on empirical methods in natural language processing, pp. 1173–1182, 2016.
2. Mohammad Gheshlaghi Azar, Zhaohan Daniel Guo, Bilal Piot, Remi Munos, Mark Rowland, Michal Valko, and Daniele Calandriello. [A general theoretical paradigm to understand learning from human preferences](https://md.research.elamur.ai/free/en/papers/azar-2023-general-theoretical-paradigm-understand-learning/). In International Conference on Artificial Intelligence and Statistics, pp. 4447–4455. PMLR, 2024. [↗](https://doi.org/10.48550/arxiv.2310.12036)
3. Kenneth Boyd. Group understanding. Synthese, 198(7):6837–6858, 2021. [find](https://scholar.google.com/scholar?q=Kenneth+Boyd%2E+Group+understanding%2E+Synthese%2C+198%287%29%3A6837%E2%80%936858%2C+2021%2E)
4. J. Butler, S. Jaffe, R. Janßen, N. Baym, B. Hecht, J. Hofman, S. Rintel, B. Sarrafzadeh, A. Sellen, M. Vorvoreanu, and J. Teevan. Microsoft new future of work report 2025. Technical Report MSR-TR-2025-58, Microsoft Research, 2025. URL [https://aka.ms/nfw2025](https://aka.ms/nfw2025). [↗](https://aka.ms/nfw2025)
5. Anabela Cantiani, Ilja van Beest, Frans Cruijssen, Goos Kant, and Thorsten M Erle. Perspective-taking predicts success in coalition formation. European Journal of Social Psychology, 54(6):1364–1377, 2024. [↗](https://doi.org/10.1002/ejsp.3091)
6. Alan Chan, Hugo Silva, Sungsu Lim, Tadashi Kozuno, A Rupam Mahmood, and Martha White. Greedification operators for policy optimization: Investigating forward and reverse kl divergences. Journal of Machine Learning Research, 23(253):1–79, 2022. [find](https://scholar.google.com/scholar?q=Alan+Chan%2C+Hugo+Silva%2C+Sungsu+Lim%2C+Tadashi+Kozuno%2C+A+Rupam+Mahmood%2C+and+Martha+White%2E+Greedification+operators+for+policy+optimization%3A+Investigating+forward+and+reverse+kl+divergences%2E+Journal+of+Mac)
7. Jiangjie Chen, Xintao Wang, Rui Xu, Siyu Yuan, Yikai Zhang, Wei Shi, Jian Xie, Shuang Li, Ruihan Yang, Tinghui Zhu, et al. [From persona to personalization: A survey on role-playing language agents](https://md.research.elamur.ai/free/en/papers/chen-2024-persona-personalization-survey-role-playing/). Transactions on Machine Learning Research, 2024.
8. Jan Chenail and Ronald Chenail. Communicating qualitative analytical results following grice’s conversational maxims. The Qualitative Report, pp. 276–285, 10 2014. doi: 10.46743/2160-3715/2011.1053. [↗](https://doi.org/10.46743/2160-3715/2011.1053)
9. Herbert H Clark. Using language. Cambridge university press, 1996. [find](https://scholar.google.com/scholar?q=Herbert+H+Clark%2E+Using+language%2E+Cambridge+university+press%2C+1996%2E)
10. Herbert H Clark and Susan E Brennan. Grounding in communication. 1991. [↗](https://doi.org/10.1037/10096-006)
11. Tim R Davidson, Adam Fourney, Saleema Amershi, Robert West, Eric Horvitz, and Ece Kamar. The collaboration gap. arXiv preprint arXiv:2511.02687, 2025. [↗](https://arxiv.org/abs/2511.02687)
12. Jacob Eisenstein, Fantine Huot, Adam Fisch, Jonathan Berant, and Mirella Lapata. Mt-pingeval: Evaluating multi-turn collaboration with private information games. arXiv preprint arXiv:2602.24188, 2026. [↗](https://arxiv.org/abs/2602.24188)
13. Lautaro Estienne, Gabriel Ben Zenou, Nona Naderi, Jackie CK Cheung, and Pablo Piantanida. Collaborative rational speech act: Pragmatic reasoning for multi-turn dialog. In Christos Christodoulopoulos, Tanmoy Chakraborty, Carolyn Rose, and Violet Peng (eds.), Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pp. 22509–22523, Suzhou, China, November 2025. Association for Computational Linguistics. ISBN 979-8-89176-332-6. doi: 10.18653/v1/2025.emnlp-main.1145. URL https://aclanthology.org/2025.emnlp-main.1145/. [↗](https://doi.org/10.18653/v1/2025.emnlp-main.1145)
14. Anna Filipi and Roger Wales. Perspective-taking and perspective-shifting as socially situated and collaborative actions. Journal of pragmatics, 36(10):1851–1884, 2004. [↗](https://doi.org/10.1016/j.pragma.2003.11.003)
15. Sammy Floyd, Olessia Jouravlev, Moshe Poliak, Zachary Mineroff, Edward Gibson, and Evelina Fedorenko. Three distinct components of pragmatic language use: Social conventions, intonation, and world knowledge–based causal reasoning. Proceedings of the National Academy of Sciences, 122(50):e2424400122, 2025. [↗](https://doi.org/10.1073/pnas.2424400122)
16. Michael C Frank and Noah D Goodman. [Predicting pragmatic reasoning in language games](https://md.research.elamur.ai/free/en/papers/frank-2012-predicting-pragmatic-reasoning-language-games/). Science, 336(6084):998–998, 2012. [↗](https://doi.org/10.1126/science.1218633)
17. Michael C Frank, Andrés Gómez Emilsson, Benjamin Peloquin, Noah D Goodman, and Christopher Potts. Rational speech act models of pragmatic reasoning in reference games. psyarxiv, 2016. [↗](https://doi.org/10.17605/osf.io/f9y6b)
18. Yujian Gan, Yuan Liang, Yanni Lin, Juntao Yu, and Massimo Poesio. Improving llms’ learning of coreference resolution. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pp. 311–321, 2025. [find](https://scholar.google.com/scholar?q=Yujian+Gan%2C+Yuan+Liang%2C+Yanni+Lin%2C+Juntao+Yu%2C+and+Massimo+Poesio%2E+Improving+llms%E2%80%99+learning+of+coreference+resolution%2E+In+Proceedings+of+the+26th+Annual+Meeting+of+the+Special+Interest+Group+on+Discour)
19. Piotr Gmytrasiewicz. How to do things with words: A bayesian approach. Journal of Artificial Intelligence Research, 68:753–776, 2020. [↗](https://doi.org/10.1613/jair.1.11951)
20. Piotr J Gmytrasiewicz and Prashant Doshi. A framework for sequential planning in multi-agent settings. Journal of Artificial Intelligence Research, 24:49–79, 2005. [↗](https://doi.org/10.1613/jair.1579)
21. Noah D Goodman and Michael C Frank. Pragmatic language interpretation as probabilistic inference. Trends in cognitive sciences, 20(11):818–829, 2016. [↗](https://doi.org/10.1016/j.tics.2016.08.005)
22. Noah D Goodman and Andreas Stuhlmüller. Knowledge and implicature: Modeling language understanding as social cognition. Topics in cognitive science, 5(1):173–184, 2013. [↗](https://doi.org/10.1111/tops.12007)
23. Herbert P Grice. Logic and conversation. In Speech acts, pp. 41–58. Brill, 1975. [↗](https://doi.org/10.1163/9789004368811_003)
24. Xudong Guo, Kaixuan Huang, Jiale Liu, Wenhui Fan, Natalia Vélez, Qingyun Wu, Huazheng Wang, Thomas L Griffiths, and Mengdi Wang. Embodied llm agents learn to cooperate in organized teams. IEEE Transactions on Computational Social Systems, 2026. [↗](https://doi.org/10.1109/tcss.2025.3637527)
25. Daniel Hardt. [Sparks of pure competence in llms: the case of syntactic center embedding in english](https://md.research.elamur.ai/free/en/papers/hardt-2025-sparks-pure-competence-llms-case/). In Proceedings of the Society for Computation in Linguistics 2025, pp. 346–355, 2025. [↗](https://doi.org/10.7275/scil.3149)
26. John Hattie. Visible learning for teachers: Maximizing impact on learning. Routledge, 2012. [↗](https://doi.org/10.4324/9780203181522)
27. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. [Lora: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/), 2021. URL https://arxiv.org/abs/2106.09685. [↗](https://arxiv.org/abs/2106.09685)
28. Jennifer Hu, Felix Sosa, and Tomer Ullman. Re-evaluating theory of mind evaluation in large language models. Philosophical Transactions of the Royal Society B: Biological Sciences, 380(1932), 2025. [↗](https://doi.org/10.1098/rstb.2023.0499)
29. Paul Jaccard. The distribution of the flora in the alpine zone. 1. New phytologist, 11(2):37–50, 1912. [find](https://scholar.google.com/scholar?q=Paul+Jaccard%2E+The+distribution+of+the+flora+in+the+alpine+zone%2E+1%2E+New+phytologist%2C+11%282%29%3A37%E2%80%9350%2C+1912%2E)
30. Mingyue Jian and N Siddharth. [Are llms good pragmatic speakers?](https://md.research.elamur.ai/free/en/papers/jian-2024-llms-good-pragmatic-speakers/) In NeurIPS 2024 Workshop on Behavioral Machine Learning, 2024. [↗](https://doi.org/10.48550/arxiv.2411.01562)
31. Daniel Jurafsky. Pragmatics and computational linguistics. The handbook of pragmatics, pp. 578–604, 2006. [↗](https://doi.org/10.1002/9780470756959.ch26)
32. Hoang Anh Just, Mahavir Dabas, Lifu Huang, Ming Jin, and Ruoxi Jia. Dipt: Enhancing llm reasoning through diversified perspective-taking. In Findings of the Association for Computational Linguistics: NAACL 2025, pp. 6344–6374, 2025. [find](https://scholar.google.com/scholar?q=Hoang+Anh+Just%2C+Mahavir+Dabas%2C+Lifu+Huang%2C+Ming+Jin%2C+and+Ruoxi+Jia%2E+Dipt%3A+Enhancing+llm+reasoning+through+diversified+perspective%2Dtaking%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+)
33. Napoleon Katsos, Blanche Gonzales de Linares, Ekaterina Ostashchenko, and Elspeth Wilson. Perspective-taking in deriving implicatures: The listener’s perspective is important too. Cognition, 241:105582, 2023. [↗](https://doi.org/10.1016/j.cognition.2023.105582)
34. Ibrahim Khalil Khebour, Kenneth Lai, Mariah Bradford, Yifan Zhu, Richard A Brutti, Christopher Tam, Jingxuan Tu, Benjamin A Ibarra, Nathaniel Blanchard, Nikhil Krishnaswamy, et al. [Common ground tracking in multimodal dialogue](https://md.research.elamur.ai/free/en/papers/khebour-2024-common-ground-tracking-multimodal-dialogue/). In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pp. 3587–3602, 2024.
35. Raihan Kibria, Sheikh Intiser Uddin Dipta, and Muhammad Abdullah Adnan. On functional competence of llms for linguistic disambiguation. In Proceedings of the 28th Conference on Computational Natural Language Learning, pp. 143–160, 2024. [find](https://scholar.google.com/scholar?q=Raihan+Kibria%2C+Sheikh+Intiser+Uddin+Dipta%2C+and+Muhammad+Abdullah+Adnan%2E+On+functional+competence+of+llms+for+linguistic+disambiguation%2E+In+Proceedings+of+the+28th+Conference+on+Computational+Natural+L)
36. Eric D Langlois and Tom Everitt. How rl agents behave when their actions are modified. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pp. 11586–11594, 2021. [↗](https://doi.org/10.1609/aaai.v35i13.17378)
37. Guohao Li, Hasan Hammoud, Hani Itani, Dmitrii Khizbullin, and Bernard Ghanem. Camel: Communicative agents for” mind” exploration of large language model society. Advances in neural information processing systems, 36:51991–52008, 2023. — [CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society](https://md.research.elamur.ai/free/en/papers/li-2023-camel-communicative-agents-mind/)
38. Bolei Ma, Yuting Li, Wei Zhou, Ziwei Gong, Yang Janet Liu, Katja Jasinskaja, Annemarie Friedrich, Julia Hirschberg, Frauke Kreuter, and Barbara Plank. [Pragmatics in the era of large language models: A survey on datasets, evaluation, opportunities and challenges](https://md.research.elamur.ai/free/en/papers/ma-2025-pragmatics-era-large-language-models/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 8679–8696, 2025a.
39. Ziqiao Ma, Jing Ding, Xuejun Zhang, Dezhi Luo, Jiahe Ding, Sihan Xu, Yuchen Huang, Run Peng, and Joyce Chai. Vision-language models are not pragmatically competent in referring expression generation. In Second Conference on Language Modeling, 2025b. [↗](https://doi.org/10.48550/arxiv.2504.16060)
40. Kiyosu Maeda, William P McCarthy, Ching-Yi Tsai, Jeffrey Mu, Haoliang Wang, Robert D Hawkins, Judith E Fan, and Parastoo Abtahi. Gesturing toward abstraction: Multimodal convention formation in collaborative physical tasks. arXiv preprint arXiv:2602.08914, 2026. [↗](https://arxiv.org/abs/2602.08914)
41. Nestor Maslej, Loredana Fattorini, Raymond Perrault, Yolanda Gil, Vanessa Parli, Njenga Kariuki, Emily Capstick, Anka Reuel, Erik Brynjolfsson, John Etchemendy, Katrina Ligett, Terah Lyons, James Manyika, Juan Carlos Niebles, Yoav Shoham, Russell Wald, Toby Walsh, Armin Hamrah, Lapo Santarlasci, Julia Betts Lotufo, Alexandra Rome, Andrew Shi, and Sukrut Oak. Artificial intelligence index report 2025, 2025. URL https://arxiv.org/abs/2504.07139. [↗](https://arxiv.org/abs/2504.07139)
42. Qingkai Min, Zitian Qu, Qipeng Guo, Xiangkun Hu, Zheng Zhang, and Yue Zhang. Multi-document event extraction using large and small language models. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pp. 19265–19296, 2025. [find](https://scholar.google.com/scholar?q=Qingkai+Min%2C+Zitian+Qu%2C+Qipeng+Guo%2C+Xiangkun+Hu%2C+Zheng+Zhang%2C+and+Yue+Zhang%2E+Multi%2Ddocument+event+extraction+using+large+and+small+language+models%2E+In+Proceedings+of+the+2025+Conference+on+Empirical+M)
43. Abhijnan Nath and Nikhil Krishnaswamy. Learning “partner-aware” collaborators in multi-party collaboration. In The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025. [find](https://scholar.google.com/scholar?q=Abhijnan+Nath+and+Nikhil+Krishnaswamy%2E+Learning+%E2%80%9Cpartner%2Daware%E2%80%9D+collaborators+in+multi%2Dparty+collaboration%2E+In+The+Thirty%2Dninth+Annual+Conference+on+Neural+Information+Processing+Systems%2C+2025%2E)
44. Abhijnan Nath, Shadi Manafi Avari, Avyakta Chelle, and Nikhil Krishnaswamy. Okay, let’s do this! modeling event coreference with generated rationales and knowledge distillation. In Proceedings of the 2024 [Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pp](https://md.research.elamur.ai/free/en/papers/nath-2024-okay-let-s-do-this/). 3931– 3946, 2024a. [↗](https://doi.org/10.48550/arxiv.2404.03196)
45. Abhijnan Nath, Videep Venkatesha, Mariah Bradford, Avyakta Chelle, Austin C Youngren, Carlos Mabrey, Nathaniel Blanchard, and Nikhil Krishnaswamy. “any other thoughts, hedgehog?” linking deliberation chains in collaborative dialogues. In Findings of the Association for Computational Linguistics: EMNLP 2024, pp. 5297–5314, 2024b. [find](https://scholar.google.com/scholar?q=Abhijnan+Nath%2C+Videep+Venkatesha%2C+Mariah+Bradford%2C+Avyakta+Chelle%2C+Austin+C+Youngren%2C+Carlos+Mabrey%2C+Nathaniel+Blanchard%2C+and+Nikhil+Krishnaswamy%2E+%E2%80%9Cany+other+thoughts%2C+hedgehog%3F%E2%80%9D+linking+deliberation+)
46. Abhijnan Nath, Carine Graff, Andrei Bachinin, and Nikhil Krishnaswamy. Frictional agent alignment framework: Slow down and don’t break things. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 11042–11089, 2025a. [find](https://scholar.google.com/scholar?q=Abhijnan+Nath%2C+Carine+Graff%2C+Andrei+Bachinin%2C+and+Nikhil+Krishnaswamy%2E+Frictional+agent+alignment+framework%3A+Slow+down+and+don%E2%80%99t+break+things%2E+In+Proceedings+of+the+63rd+Annual+Meeting+of+the+Associat)
47. Abhijnan Nath, Carine Graff, and Nikhil Krishnaswamy. Collaborate, deliberate, evaluate: How llm alignment affects coordinated multi-agent outcomes. arXiv preprint arXiv:2509.05882, 2025b. [↗](https://arxiv.org/abs/2509.05882)
48. Abhijnan Nath, Carine Graff, and Nikhil Krishnaswamy. Let’s roleplay: Examining llm alignment in collaborative dialogues. arXiv e-prints, pp. arXiv–2509, 2025c. [find](https://scholar.google.com/scholar?q=Abhijnan+Nath%2C+Carine+Graff%2C+and+Nikhil+Krishnaswamy%2E+Let%E2%80%99s+roleplay%3A+Examining+llm+alignment+in+collaborative+dialogues%2E+arXiv+e%2Dprints%2C+pp%2E+arXiv%E2%80%932509%2C+2025c%2E)
49. Abhijnan Nath, Hannah VanderHoeven, and Nikhil Krishnaswamy. Craft: Grounded multi-agent coordination under partial information, 2026. URL https://arxiv.org/abs/2603.25268. [↗](https://arxiv.org/abs/2603.25268)
50. Khanh Xuan Nguyen. Language models are bounded pragmatic speakers. In First Workshop on Theory of Mind in Communicating Agents, 2023. [find](https://scholar.google.com/scholar?q=Khanh+Xuan+Nguyen%2E+Language+models+are+bounded+pragmatic+speakers%2E+In+First+Workshop+on+Theory+of+Mind+in+Communicating+Agents%2C+2023%2E)
51. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. [Training language models to follow instructions with human feedback](https://md.research.elamur.ai/free/en/papers/ouyang-2022-training-language-models-follow-instructions/). Advances in neural information processing systems, 35:27730–27744, 2022.
52. Eric Pacuit. Neighborhood semantics for modal logic. Springer, 2017. [↗](https://doi.org/10.1007/978-3-319-67149-9)
53. Dojun Park, Jiwoo Lee, Seohyun Park, Hyeyun Jeong, Youngeun Koo, Soonha Hwang, Seonwoo Park, and Sungeun Lee. [Multiprageval: Multilingual pragmatic evaluation of large language models](https://md.research.elamur.ai/free/en/papers/park-2024-multiprageval-multilingual-pragmatic-evaluation-large/). In Proceedings of the 2nd GenBench Workshop on Generalisation (Benchmarking) in NLP, pp. 96–119, 2024.
54. David V Perkins and Michael J Tagler. Jigsaw classroom. Promoting student engagement, 1: 195–197, 2011. [find](https://scholar.google.com/scholar?q=David+V+Perkins+and+Michael+J+Tagler%2E+Jigsaw+classroom%2E+Promoting+student+engagement%2C+1%3A+195%E2%80%93197%2C+2011%2E)
55. Michele Poradzisz and Kristine L Florczak. Collaboration: Does it require pragmatic thought? Nursing Science Quarterly, 32(4):271–275, 2019. [↗](https://doi.org/10.1177/0894318419864336)
56. David Premack and Guy Woodruff. Does the chimpanzee have a theory of mind? Behavioral and brain sciences, 1(4):515–526, 1978. [↗](https://doi.org/10.1017/s0140525x00076512)
57. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. [Direct preference optimization: Your language model is secretly a reward model](https://md.research.elamur.ai/free/en/papers/rafailov-2023-direct-preference-optimization-language-model/). Advances in neural information processing systems, 36:53728–53741, 2024. [↗](https://doi.org/10.48550/arxiv.2305.18290)
58. Nils Reimers and Iryna Gurevych. [Sentence-bert: Sentence embeddings using siamese bert-networks](https://md.research.elamur.ai/free/en/papers/reimers-2019-sentence-bert-sentence-embeddings-siamese/). In Proceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP), pp. 3982–3992, 2019.
59. Ayse Pinar Saygin and Ilyas Cicekli. Pragmatics in human-computer conversations. Journal of Pragmatics, 34(3):227–258, 2002. [find](https://scholar.google.com/scholar?q=Ayse+Pinar+Saygin+and+Ilyas+Cicekli%2E+Pragmatics+in+human%2Dcomputer+conversations%2E+Journal+of+Pragmatics%2C+34%283%29%3A227%E2%80%93258%2C+2002%2E)
60. Claude Elwood Shannon. A mathematical theory of communication. The Bell system technical journal, 27(3):379–423, 1948. [↗](https://doi.org/10.1002/j.1538-7305.1948.tb01338.x)
61. Settaluri Sravanthi, Meet Doshi, Pavan Tankala, Rudra Murthy, Raj Dabre, and Pushpak Bhattacharyya. Pub: A pragmatics understanding benchmark for assessing llms’ pragmatics capabilities. In Findings of the Association for Computational Linguistics: ACL 2024, pp. 12075–12097, 2024. — [PUB: A Pragmatics Understanding Benchmark for Assessing LLMs' Pragmatics Capabilities](https://md.research.elamur.ai/free/en/papers/sravanthi-2024-pub-pragmatics-understanding-benchmark-assessing/)
62. Tomer Ullman. Large language models fail on trivial alterations to theory-of-mind tasks. arXiv preprint arXiv:2302.08399, 2023. [↗](https://arxiv.org/abs/2302.08399)
63. Johan Van Benthem and Eric Pacuit. Dynamic logics of evidence-based beliefs. Studia Logica, 99(1):61, 2011. [↗](https://doi.org/10.1007/s11225-011-9347-x)
64. Johan Van Benthem, David Fernández-Duque, and Eric Pacuit. Evidence and plausibility in neighborhood structures. Annals of Pure and Applied Logic, 165(1):106–133, 2014. [↗](https://doi.org/10.1016/j.apal.2013.07.007)
65. Robert Van Rooy. Utility, informativity and protocols. Journal of philosophical logic, 33(4): 389–419, 2004. [find](https://scholar.google.com/scholar?q=Robert+Van+Rooy%2E+Utility%2C+informativity+and+protocols%2E+Journal+of+philosophical+logic%2C+33%284%29%3A+389%E2%80%93419%2C+2004%2E)
66. Francis Ward, Francesca Toni, Francesco Belardinelli, and Tom Everitt. Honesty is the best policy: defining and mitigating ai deception. Advances in neural information processing systems, 36:2313–2341, 2023. [find](https://scholar.google.com/scholar?q=Francis+Ward%2C+Francesca+Toni%2C+Francesco+Belardinelli%2C+and+Tom+Everitt%2E+Honesty+is+the+best+policy%3A+defining+and+mitigating+ai+deception%2E+Advances+in+neural+information+processing+systems%2C+36%3A2313%E2%80%932341)
67. Alex Wilf, Sihyun Lee, Paul Pu Liang, and Louis-Philippe Morency. Think twice: Perspective-taking improves large language models’ theory-of-mind capabilities. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 8292[–8308, 2024](https://md.research.elamur.ai/free/en/papers/wilf-2024-think-twice-perspective-taking-improves/).
68. Guande Wu, Chen Zhao, Claudio Silva, and He He. Your co-workers matter: Evaluating collaborative capabilities of language models in blocks world, 2024. URL https://arxiv.org/abs/2404.00246. [↗](https://arxiv.org/abs/2404.00246)
69. Sina Zarrieß and David Schlangen. Know what you don’t know: Modeling a pragmatic speaker that refers to objects of unknown categories. In Proceedings of the 57th annual meeting of the association for computational linguistics, pp. 654–659, 2019. — [Know What You Don't Know: Modeling a Pragmatic Speaker that Refers to Objects of Unknown Categories](https://md.research.elamur.ai/free/en/papers/zarriess-2019-know-what-you-don-t/)
70. Lixing Zhu, Jun Wang, and Yulan He. Llmlink: Dual llms for dynamic entity linking on long narratives with collaborative memorisation and prompt optimisation. In Proceedings of the 31st international conference on computational linguistics, pp. 11334–11347, 2025a. [find](https://scholar.google.com/scholar?q=Lixing+Zhu%2C+Jun+Wang%2C+and+Yulan+He%2E+Llmlink%3A+Dual+llms+for+dynamic+entity+linking+on+long+narratives+with+collaborative+memorisation+and+prompt+optimisation%2E+In+Proceedings+of+the+31st+international+c)
71. Yifan Zhu, Changsoo Jung, Kenneth Lai, Videep Venkatesha, Mariah Bradford, Jack Fitzgerald, Huma Jamil, Carine Graff, Sai Kiran Ganesh Kumar, Bruce Draper, et al. Multimodal common ground annotation for partial information collaborative problem solving. In Proceedings of the 21st Joint ACL-ISO Workshop on Interoperable Semantic Annotation (ISA-21), pp. 85–91, 2025b. [find](https://scholar.google.com/scholar?q=Yifan+Zhu%2C+Changsoo+Jung%2C+Kenneth+Lai%2C+Videep+Venkatesha%2C+Mariah+Bradford%2C+Jack+Fitzgerald%2C+Huma+Jamil%2C+Carine+Graff%2C+Sai+Kiran+Ganesh+Kumar%2C+Bruce+Draper%2C+et+al%2E+Multimodal+common+ground+annotation+f)
72. Yifan Zhu, Mariah Bradford, Kenneth Lai, Timothy Obiso, Videep Venkatesha, James Pustejovsky, and Nikhil Krishnaswamy. Distributed partial information puzzles: Examining common ground construction under epistemic asymmetry. arXiv preprint arXiv:2603.05450, 2026. [↗](https://arxiv.org/abs/2603.05450)

*Updated 1 November 2026*