CalBench: Evaluating Coordination-Privacy Trade-offs in Multi-Agent LLMs

We introduce CalBench, a controlled evaluation environment for studying multi-agent coordination through calendar scheduling. In CalBench, N agents each manage a private calendar containing pre-existing commitments and must coordinate to schedule a stream of M incoming meetings while minimizing disruption costs. Because agents observe only their own calendars, successful scheduling requires communication across private information boundaries. Each scenario is generated with an oracle solution, enabling precise measurement of coordination quality via realized-to-optimal cost, as well as a Distributed Constraint Optimization (DCOP) baseline to provide a fair comparison under the same private-information constraints. CalBench enables precise verification of task success, communication efficiency, and fairness in the distribution of disruption costs. Our environment also studies privacy-preserving coordination by augmenting calendar entries with private semantic contexts of varying sensitivity and measuring whether agents reveal task-irrelevant private information during negotiation. Unlike multi-agent benchmarks where a single capable agent can often substitute for the group, CalBench is inherently decentralized: no agent has access to another agent's private calendar, yet agents must still reach mutually consistent decisions over shared meeting scheduling. CalBench therefore provides a practical and verifiable setting for studying coordination protocols, communication efficiency, negotiation strategies, fairness, and privacy leakage in multi-agent systems.

Talk is Cheap, Communication is Hard: Dynamic Grounding Failures and Repair in Multi-Agent Negotiation

Grounding is the collaborative process of establishing mutual belief sufficient for the current communicative purpose. While static grounding maps language to a shared, externally observable context, dynamic grounding is a joint activity where meaning is negotiated through interaction. Current multi-agent Large Language Model (LLM) benchmarks focus on static, one-shot tasks, overlooking the ability to repair grounding breakdowns across turns. We introduce an iterated, multi-turn negotiation game in which two agents allocate shared resources toward private projects with verifiable jointly optimal outcomes. While individual agents can identify Pareto-optimal allocations in isolation, agent dyads consistently fail to reach them across open- and closed-source models. Our investigation reveals four failure modes: (1) coordination degrades when shared interaction history is absent; (2) yet accumulated context can itself become a liability through stubborn anchoring, where initial proposals are treated as axiomatic rather than negotiable; (3) a reliance on perfunctory fairness (equal resource splits) over reward-maximizing coordination; and (4) failures in referential binding, where agents lose track of commitments across turns. These results highlight dynamic grounding as a critical and understudied axis of multi-agent coordination. Our framework decomposes the coordination gap into measurable components: the oracle baseline establishes that the gap is not attributable to individual reasoning limitations; the no-talk baseline establishes that communication is necessary; and a full-transparency intervention establishes that information exchange alone is insufficient: the bottleneck lies in the interactive processes of joint plan formation, commitment, and execution that constitute dynamic grounding.

Gesturing Toward Abstraction: Multimodal Convention Formation in Collaborative Physical Tasks

A quintessential feature of human intelligence is the ability to create ad hoc conventions over time to achieve shared goals efficiently. We investigate how communication strategies evolve through repeated collaboration as people coordinate on shared procedural abstractions. To this end, we conducted an online unimodal study (n = 98) using natural language to probe abstraction hierarchies. In a follow-up lab study (n = 40), we examined how multimodal communication (speech and gestures) changed during physical collaboration. Pairs used augmented reality to isolate their partner's hand and voice; one participant viewed a 3D virtual tower and sent instructions to the other, who built the physical tower. Participants became faster and more accurate by establishing linguistic and gestural abstractions and using cross-modal redundancy to emphasize key changes from previous interactions. Based on these findings, we extend probabilistic models of convention formation to multimodal settings, capturing shifts in modality preferences. Our findings and model provide building blocks for designing convention-aware intelligent agents situated in the physical world.

Act or Clarify? Modeling Sensitivity to Uncertainty and Cost in Communication

When deciding how to act under uncertainty, agents may choose to act to reduce uncertainty or they may act despite that uncertainty.In communicative settings, an important way of reducing uncertainty is by asking clarification questions (CQs). We predict that the decision to ask a CQ depends on both contextual uncertainty and the cost of alternative actions, and that these factors interact: uncertainty should matter most when acting incorrectly is costly. We formalize this interaction in a computational model based on expected regret: how much an agent stands to lose by acting now rather than with full information. We test these predictions in two experiments, one examining purely linguistic responses to questions and another extending to choices between clarification and non-linguistic action. Taken together, our results suggest a rational tradeoff: humans tend to seek clarification proportional to the risk of substantial loss when acting under uncertainty.

Accommodation and Epistemic Vigilance: A Pragmatic Account of Why LLMs Fail to Challenge Harmful Beliefs

Large language models (LLMs) frequently fail to challenge users' harmful beliefs in domains ranging from medical advice to social reasoning. We argue that these failures can be understood and addressed pragmatically as consequences of LLMs defaulting to accommodating users' assumptions and exhibiting insufficient epistemic vigilance. We show that social and linguistic factors known to influence accommodation in humans (at-issueness, linguistic encoding, and source reliability) similarly affect accommodation in LLMs, explaining performance differences across three safety benchmarks that test models' ability to challenge harmful beliefs, spanning misinformation (Cancer-Myth, SAGE-Eval) and sycophancy (ELEPHANT). We further show that simple pragmatic interventions, such as adding the phrase "wait a minute", significantly improve performance on these benchmarks while preserving low false-positive rates. Our results highlight the importance of considering pragmatics for evaluating LLM behavior and improving LLM safety.

Minding the Politeness Gap in Cross-cultural Communication

Misunderstandings in cross-cultural communication often arise from subtle differences in interpretation, but it is unclear whether these differences arise from the literal meanings assigned to words or from more general pragmatic factors such as norms around politeness and brevity. In this paper, we report three experiments examining how speakers of British and American English interpret intensifiers like "quite" and "very." To better understand these cross-cultural differences, we developed a computational cognitive model where listeners recursively reason about speakers who balance informativity, politeness, and utterance cost. Our model comparisons suggested that cross-cultural differences in intensifier interpretation stem from a combination of (1) different literal meanings, (2) different weights on utterance cost. These findings challenge accounts based purely on semantic variation or politeness norms, demonstrating that cross-cultural differences in interpretation emerge from an intricate interplay between the two.

Comparing human and LLM politeness strategies in free production

Polite speech poses a fundamental alignment challenge for large language models (LLMs). Humans deploy a rich repertoire of linguistic strategies to balance informational and social goals -- from positive approaches that build rapport (compliments, expressions of interest) to negative strategies that minimize imposition (hedging, indirectness). We investigate whether LLMs employ a similarly context-sensitive repertoire by comparing human and LLM responses in both constrained and open-ended production tasks. We find that larger models ($\ge$70B parameters) successfully replicate key preferences from the computational pragmatics literature, and human evaluators surprisingly prefer LLM-generated responses in open-ended contexts. However, further linguistic analyses reveal that models disproportionately rely on negative politeness strategies even in positive contexts, potentially leading to misinterpretations. While modern LLMs demonstrate an impressive handle on politeness strategies, these subtle differences raise important questions about pragmatic alignment in AI systems.

Learning a Hierarchical Planner from Humans in Multiple Generations

A typical way in which a machine acquires knowledge from humans is by programming. Compared to learning from demonstrations or experiences, programmatic learning allows the machine to acquire a novel skill as soon as the program is written, and, by building a library of programs, a machine can quickly learn how to perform complex tasks. However, as programs often take their execution contexts for granted, they are brittle when the contexts change, making it difficult to adapt complex programs to new contexts. We present natural programming, a library learning system that combines programmatic learning with a hierarchical planner. Natural programming maintains a library of decompositions, consisting of a goal, a linguistic description of how this goal decompose into sub-goals, and a concrete instance of its decomposition into sub-goals. A user teaches the system via curriculum building, by identifying a challenging yet not impossible goal along with linguistic hints on how this goal may be decomposed into sub-goals. The system solves for the goal via hierarchical planning, using the linguistic hints to guide its probability distribution in proposing the right plans. The system learns from this interaction by adding newly found decompositions in the successful search into its library. Simulated studies and a human experiment (n=360) on a controlled environment demonstrate that natural programming can robustly compose programs learned from different users and contexts, adapting faster and solving more complex tasks when compared to programmatic baselines.

Causal interventions expose implicit situation models for commonsense language understanding

Accounts of human language processing have long appealed to implicit ``situation models'' that enrich comprehension with relevant but unstated world knowledge. Here, we apply causal intervention techniques to recent transformer models to analyze performance on the Winograd Schema Challenge (WSC), where a single context cue shifts interpretation of an ambiguous pronoun. We identify a relatively small circuit of attention heads that are responsible for propagating information from the context word that guides which of the candidate noun phrases the pronoun ultimately attends to. We then compare how this circuit behaves in a closely matched ``syntactic'' control where the situation model is not strictly necessary. These analyses suggest distinct pathways through which implicit situation models are constructed to guide pronoun resolution.

Semantic uncertainty guides the extension of conventions to new referents

A long tradition of studies in psycholinguistics has examined the formation and generalization of ad hoc conventions in reference games, showing how newly acquired conventions for a given target transfer to new referential contexts. However, another axis of generalization remains understudied: how do conventions formed for one target transfer to completely distinct targets, when specific lexical choices are unlikely to repeat? This paper presents two dyadic studies (N = 240) that address this axis of generalization, focusing on the role of nameability -- the a priori likelihood that two individuals will share the same label. We leverage the recently-released KiloGram dataset, a collection of abstract tangram images that is orders of magnitude larger than previously available, exhibiting high diversity of properties like nameability. Our first study asks how nameability shapes convention formation, while the second asks how new conventions generalize to entirely new targets of reference. Our results raise new questions about how ad hoc conventions extend beyond target-specific re-use of specific lexical choices.