Yang, Xiulin, Gudavalli, Sirisha, Hu, Jennifer · Nov 10, 2026
Rational Use of Time for Social Coordination in Humans and AI
Markdown · JSON · Original source
Xiulin Yang, Sirisha Gudavalli and Jennifer Hu study whether humans and AI can use timing to coordinate when they cannot communicate explicitly. In The Mind, a cooperative card game, the passage of is the only available signal. The authors evaluate 150 humans alongside computer-use agents (CUAs) and standard text-prompted . They report that a Bayesian strategy—in which a player infers a partner’s belief from their timing—best captures human behavior and, similarly, . But they draw a distinction between inferring and adapting: none of the tested models seem to efficiently adjust their behavior to their partners over the course of the game. The authors interpret the human results as evidence of implicit probabilistic reasoning about continuous-time signals in dynamic social settings, and suggest that such reasoning may emerge without real-time experience. For AI, the practical challenge they highlight is not just reading an implicit signal, but learning to adapt to it while coordinating with another agent. The abstract does not establish how well these findings extend beyond this game or which changes would help models adapt.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
cs.AI · Article
Source details
Original source date: Nov 10, 2026 (2026-11-10).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-13T11:33:33.457Z. First observed: 2026-11-13T07:42:51.168Z. Retrieved: 2026-11-13T11:32:39.540Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
The supplied evidence is the paper’s abstract, not its full methods or results; details needed to assess the evaluation are not available here. The reported lack of efficient adaptation is limited to the tested models and the game described in the abstract. The abstract does not establish how the findings generalize beyond The Mind or what changes would improve model adaptation.
Open question
What would enable the tested models to adapt their behavior efficiently to their partners over the course of the game?
Source-stated
The study uses The Mind, a cooperative card game in which players cannot communicate explicitly and time is the only signal.
the only available signal is the passage of time
Source-stated
The authors evaluate 150 humans, computer-use agents (CUAs), and standard text-prompted LLMs.
We evaluate 150 humans, and two classes of AI models
Source-stated
The authors report that a Bayesian strategy based on inferring a partner’s belief from timing best captures human behavior, and similarly captures frontier models.
human behavior is best captured by a Bayesian strategy that infers a partner's belief from their timing---and similarly for frontier models.
Source-stated
The authors report that none of the tested models seem to efficiently adapt their behavior to partners during the game.
none of the tested models seem to efficiently adapt their behaviors to their partners' over the course of the game.
Source-stated
The authors suggest that humans perform implicit probabilistic reasoning about continuous-time signals, and that these reasoning mechanisms may emerge without real-time experience.
such reasoning mechanisms may emerge without real-time experience.
Source and contributor context
Yang, Xiulin, Gudavalli, Sirisha, Hu, Jennifer · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Frontier model
One of the most capable AI models available at a given time; the set shifts as models improve.
TIME
TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: cbb66c34-6e40-4661-86cb-1ab3cd79b1e2.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: passed by anthropic/claude-opus-5.5 on 2026-11-13T11:33:33.457Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: 85f1c1c896d1866f02653efcf1a72f9b283d35854e656decda4080f4937fa8e4. Source text hash: e25b6d41e86ad5e2fe6a7c6ce1880f1b36be14e9b32b2668900f7827104dbe81.
Revision and source-check history
Version 1 · 2026-11-13T11:33:33.457Z
Rational Use of Time for Social Coordination in Humans and AI
Xiulin Yang, Sirisha Gudavalli and Jennifer Hu study whether humans and AI can use timing to coordinate when they cannot communicate explicitly. In The Mind, a cooperative card game, the passage of is the only available signal. The authors evaluate 150 humans alongside computer-use agents (CUAs) and standard text-prompted . They report that a Bayesian strategy—in which a player infers a partner’s belief from their timing—best captures human behavior and, similarly, . But they draw a distinction between inferring and adapting: none of the tested models seem to efficiently adjust their behavior to their partners over the course of the game. The authors interpret the human results as evidence of implicit probabilistic reasoning about continuous-time signals in dynamic social settings, and suggest that such reasoning may emerge without real-time experience. For AI, the practical challenge they highlight is not just reading an implicit signal, but learning to adapt to it while coordinating with another agent. The abstract does not establish how well these findings extend beyond this game or which changes would help models adapt.
Original source
cs.AI · Article
Source details
Original source date: Nov 10, 2026 (2026-11-10).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-13T11:33:33.457Z. First observed: 2026-11-13T07:42:51.168Z. Retrieved: 2026-11-13T11:32:39.540Z.
Limits
The supplied evidence is the paper’s abstract, not its full methods or results; details needed to assess the evaluation are not available here. The reported lack of efficient adaptation is limited to the tested models and the game described in the abstract. The abstract does not establish how the findings generalize beyond The Mind or what changes would improve model adaptation.
Open question
What would enable the tested models to adapt their behavior efficiently to their partners over the course of the game?
Source-stated
The study uses The Mind, a cooperative card game in which players cannot communicate explicitly and time is the only signal.
the only available signal is the passage of time
Source-stated
The authors evaluate 150 humans, computer-use agents (CUAs), and standard text-prompted LLMs.
We evaluate 150 humans, and two classes of AI models
Source-stated
The authors report that a Bayesian strategy based on inferring a partner’s belief from timing best captures human behavior, and similarly captures frontier models.
human behavior is best captured by a Bayesian strategy that infers a partner's belief from their timing---and similarly for frontier models.
Source-stated
The authors report that none of the tested models seem to efficiently adapt their behavior to partners during the game.
none of the tested models seem to efficiently adapt their behaviors to their partners' over the course of the game.
Source-stated
The authors suggest that humans perform implicit probabilistic reasoning about continuous-time signals, and that these reasoning mechanisms may emerge without real-time experience.
such reasoning mechanisms may emerge without real-time experience.
Source and contributor context
Yang, Xiulin, Gudavalli, Sirisha, Hu, Jennifer · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Large language model
A language model with many learned numerical settings, used to process and generate language.
Frontier model
One of the most capable AI models available at a given time; the set shifts as models improve.
TIME
TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: cbb66c34-6e40-4661-86cb-1ab3cd79b1e2.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: passed by anthropic/claude-opus-5.5 on 2026-11-13T11:33:33.457Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: 85f1c1c896d1866f02653efcf1a72f9b283d35854e656decda4080f4937fa8e4. Source text hash: e25b6d41e86ad5e2fe6a7c6ce1880f1b36be14e9b32b2668900f7827104dbe81.
passed · 2026-11-13T11:33:33.457Z · anthropic/claude-opus-5.5
Publication: cbb66c34-6e40-4661-86cb-1ab3cd79b1e2
attribution — pass: The reading correctly credits the study to Xiulin Yang, Sirisha Gudavalli and Jennifer Hu. Findings and interpretations are framed as the authors' reports and suggestions, not as established fact or EM opinion.
Authors:Xiulin Yang, Sirisha Gudavalli, Jennifer Hu
summary — pass: The title matches the source. The summary explains The Mind's key constraint, spells out CUAs and LLMs, and defines the Bayesian strategy in plain terms. It keeps the authors' inference-versus-adaptation contrast and their hedged cognitive claim. The game's goal is not described, but the abstract does not give it either, and the central idea is still clear. The question is open and follows from the reported adaptation gap.
none of the tested models seem to efficiently adapt their behaviors to their partners' over the course of the game
date — pass: The originalPublished date of 2026-11-10 matches the arXiv v1 submission date. The later feed date (2026-11-13) was correctly not used.
[Submitted on 10 Nov 2026]
support — pass: Each claim matches the abstract and keeps its hedging, including 'seem to' and 'may emerge'. The practical challenge mirrors the authors' wording about learning to interpret and adapt to implicit signals. The limitations correctly note that the evidence is an abstract only and that generalization is unknown. No benefits, costs or forecasts are invented.
our study highlights new challenges for AI systems to coordinate efficiently with other agents by learning to interpret and adapt to their implicit signals in dynamic settings
media — pass: The selected media is the verified direct arXiv abstract page for this paper, which is the most direct way to reach the source.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
Explore TIMEA language model with many learned numerical settings, used to process and generate language.
Definition referenceOne of the most capable AI models available at a given time; the set shifts as models improve.
Definition referencearXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivTIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
Explore TIMEA language model with many learned numerical settings, used to process and generate language.
Definition referenceOne of the most capable AI models available at a given time; the set shifts as models improve.
Definition referencearXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXiv