Pith. sign in

REVIEW 3 major objections 4 minor 75 references

Conversational Alignment with Artificial Intelligence in Context

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read The paper proposes an 11-point framework for conversational alignment and argues that current LLMs cannot fully satisfy it because retaining context and separating contexts pull in opposite directions.

desk verdict Useful synthesis of pragmatics and AI alignment, but the 'fundamental limitations' claim outruns the argument; the CONTEXT-ALIGN framework is the real contribution. read the letter →

arxiv 2505.22907 v1 pith:ERYVHBDZ submitted 2025-05-28 cs.CY cs.CL

classification cs.CYcs.CL
keywords conversationalalignmentcontextcollapselargelanguagemodelscommongroundpragmaticswindowoverflowhuman-AIagents
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks what it would mean for an AI assistant to be conversationally aligned with human communicative practices, and proposes an 11-point framework called CONTEXT-ALIGN for evaluating that. It argues that current large language models fail most of these criteria, and that the failures are not purely matters of training data or prompt quality: the finite context window forces a choice between forgetting earlier parts of a conversation and blending distinct conversational contexts together, while prompting and fixed alignment personas replace the jointly built common ground of human dialogue. The stakes are practical as well as philosophical: if the argument is right, conversational breakdowns, misplaced trust, and miscommunication in high-stakes AI deployments are built into the architecture, and achieving full conversational alignment requires rethinking the models themselves, not just tuning them.

What carries the argument

The load-bearing object is the CONTEXT-ALIGN framework, a set of eleven desiderata—tracking context-sensitive meaning, common-ground management, conversational-scoreboard updating, discourse and QUD management, accommodation, pragmatic inference, ethical-pragmatic integration, context-collapse mitigation, repair protocols, transparency, and cross-contextual memory—that the paper uses as a yardstick. The mechanism that carries the negative claim is the context-window overflow/context-collapse dilemma: retaining more history to preserve coherence pushes the model toward merging distinct contexts, while compressing or dropping history sacrifices coherence. A second mechanism is the prompt-as-context substitution, in which the user's static prompt must stand in for the dynamic common ground that human interlocutors co-construct.

What would settle it

A controlled evaluation would settle the core tension: take a long multi-topic conversation in which a user switches from a software bug to a billing query, and compare an assistant that stores all history as one flat context against one that keeps per-thread context metadata. If the structured-memory assistant maintains both coherence and correct separation across, say, fifty turns at rates comparable to human judges, while the flat-context assistant does not, the claimed fundamental trade-off would be shown to be an artifact of current memory design rather than a limit on conversational alignment itself.

Watch

Extended reading notes

Core claim

The paper's central claim is that LLMs 'fall well short' of the desiderata for full conversational alignment, and that the gap may be fundamental rather than contingent. Its argument has two load-bearing parts. First, because an LLM manages an entire conversation as a single string of tokens inside a finite window, any workable memory strategy—summarizing, enlarging the window, or retrieving from external storage—either loses old context or collapses distinct conversational threads into one frame, so there is no way to simultaneously satisfy coherence and context-separation. Second, the pragmatics of prompting force users to supply static context unilaterally, while behavioral alignment protocols such as HHH impose a fixed communicative identity on the model, leaving it unable to negotiate norms, repair misunderstanding, or adapt its persona the way human speakers do. The conclusion is that current transformer-based agents are ethically constrained but pragmatically alien conversational partners.

Load-bearing premise

The load-bearing premise is that an LLM manages context as a single undifferentiated token string rather than a structured model of separate threads and audiences, so that any memory strategy must either lose old context or blend distinct contexts together.

Editorial extensions

If this is right

  • For present-day transformer-based assistants, satisfying all 11 criteria at once is not attainable by prompt engineering or larger context windows, because any increase in remembered context raises the risk of blending distinct conversational threads.
  • The burden of building and maintaining shared context falls almost entirely on the user, who must pre-scaffold the model's role, audience, register, and goals in the prompt rather than negotiating them over the course of dialogue.
  • Ethical alignment as currently practiced can actively work against conversational alignment: a model trained to be honest or neutral may refuse to simplify, joke, or tailor its persona even when the local context calls for it.
  • Moving toward the framework's goals requires architectural changes such as explicit context-negotiation, context-aware norm weighting, user-driven persona customization, and letting models propose norm exceptions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable corollary the paper does not spell out: the context-overflow/context-collapse trade-off should show up as an inverted-U curve in evaluation, where pragmatic appropriateness rises then falls as memory length grows; a benchmark could measure that directly.
  • If the limitation is architectural, then hybrid systems that add structured discourse-state tracking (separate per-thread QUD stacks, audience-tagged memory) are the natural intervention; their success or failure would decide whether the claimed tension is truly fundamental.
  • The social-media context-collapse literature suggests a design target: LLMs could be given an explicit notion of imagined audience so that they can separate contexts the way humans do; the paper gestures at this but does not develop it.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper develops a conceptual framework, CONTEXT-ALIGN, with eleven desiderata drawn from philosophical pragmatics and the philosophy of language, covering context-sensitivity, common ground, conversational scoreboards, discourse structure, accommodation, pragmatic inference, ethical-pragmatic integration, context-collapse mitigation, repair protocols, transparency, and cross-contextual memory. It then argues that LLM-based conversational agents fall well short of these desiderata. Section 4 claims that context-window overflow creates a fundamental tension: retaining too much context risks context collapse, while retaining too little risks incoherence and loss of common ground. Section 5 argues that prompting is a poor substitute for dynamically co-constructed context and that behavioral alignment protocols impose a static communicative identity on LLMs. Section 6 concludes that current LLM architectures may be fundamentally limited in achieving full conversational alignment, while hedging that the limitations might not be insurmountable.

Significance. If established, the paper would make a valuable contribution by separating conversational alignment from ethical alignment and by providing a structured checklist, the CONTEXT-ALIGN criteria, for evaluating AI interlocutors. The framework is well grounded in a broad philosophical literature, and the authors are appropriately cautious about the strength of their concluding claim. The main value is currently in framing and taxonomy rather than in proving an impossibility result: the strongest negative thesis depends on an empirical premise about how LLMs represent and store context that is asserted with illustrative examples rather than demonstrated. The paper does not ship machine-checked proofs, code, or falsifiable experiments, so its significance rests on the plausibility of its conceptual analysis and on whether the load-bearing empirical premise survives scrutiny.

major comments (3)
  1. [§4.2] The load-bearing premise is the statement that 'LLMs do not maintain a structured, human-like model of the conversation but instead manage context as a single long string of tokens from previous turns.' This premise is used to infer that any memory-aiding strategy (compression, larger windows, external memory, RAG) necessarily risks context collapse. The inference does not follow: the fact that the input is a flat token sequence does not imply that the model's internal state or an external memory system cannot separate conversational threads. Attention and hidden states can in principle encode discourse structure, and external memory with thread IDs, timestamps, or topic tags could segregate contexts; the paper gives no argument that such mechanisms are absent or impossible in current architectures. The anecdotal examples (software bug vs. billing; philosophy vs. vacation) show only that a flat-history implementation can blend contexts, not that current architectures must do so. Moreover, §4.3 itself calls for 'structured context management,' which weakens the claim that the tension is architectural rather than implementational. The paper should either restrict its claim to standard single-history implementations or provide a substantive argument that structured memory and metadata cannot prevent context collapse; without this, the 'fundamental tension' of §4.2 and the 'insurmountable in principle' hedge of §6 are unsupported.
  2. [§5.2] The claim that 'persuading an LLM to adopt a sarcastic tone is not possible without jailbreaking' is not supported and is arguably false for many current instruction-tuned models, which will adopt a specified style, persona, or register when asked in the system prompt or through in-context examples. This example is used to illustrate the 'static communicative identity' thesis and to support the violation of CA8; without it, the thesis must be narrowed to value-level alignment under safety training rather than to tone or style, or it needs empirical support. Please either revise the example, distinguish 'cannot be changed through dialogue' from 'cannot be changed through ordinary dialogue under default settings,' and provide evidence for the narrower claim.
  3. [§5.1] The claim that 'LLMs treat prompts as static contextual frames' and that 'subsequent turns are interpreted strictly within this frame' is asserted without empirical or architectural evidence. Modern instruction-tuned models can adapt to mid-conversation topic shifts and can re-scaffold context in response to follow-up user corrections, and the paper's own proposed mitigations (querying user goals, flagging ambiguous cues, context verification) are precisely mechanisms that would allow such adaptation. As written, this section overstates the staticity of prompt-based context and needs either experimental support or a more careful restriction to particular system configurations.
minor comments (4)
  1. [§2.4] The phrase 'perineal meteorological ignorance' appears to be a typo for 'perennial meteorological ignorance.'
  2. [§4.2] The sentence 'The resulting phenomena is similar to context collapse in social media' should read 'The resulting phenomenon is similar' (or 'phenomena are similar'), and the relation between social-media context collapse and LLM topic-blending deserves a footnote explaining the intended analogy and its limits.
  3. [Abstract and §6] The paper oscillates between 'may impose fundamental limitations' (abstract), 'fundamental tension' (§4.2), and 'insurmountable in principle, although they might not be' (§6). Please state precisely what modal and scope claims are being made, e.g., 'for current Transformer-based agents with standard training pipelines' versus 'for any possible implementation of the criterion list.'
  4. [§3, criterion CA5] CA5 says the model should accommodate repairs and clarifications 'without requiring restatements of prior information,' but CA9 and the mitigation proposals in §5.1 explicitly invite users to query and re-state contextual information. The tension between these should be acknowledged and resolved.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the CONTEXT-ALIGN desiderata are imported from external philosophy and linguistics sources, and the claim that LLMs fall short is an application of those independent criteria, not a conclusion defined into the premises.

full rationale

The paper's derivation chain is not circular. The CONTEXT-ALIGN criteria are drawn from external sources (Stalnaker, Lewis, Grice, Roberts, Meyrowitz, boyd, and others) and are not defined in terms of LLM behavior; the paper then assesses LLMs against those independently motivated desiderata. The central negative claim in Section 6 does not reduce to the framework's definition: it depends on an empirical and architectural premise, stated in Section 4.2 ('LLMs do not maintain a structured, human-like model of the conversation but instead manage context as a single long string of tokens'), which is unsupported in the paper but is not equivalent to the conclusion. The only self-citation is Sterken (2015) in Section 2.1, used as one illustration that context-sensitivity extends to generics; it is peripheral and not load-bearing. The paper also explicitly hedges the strong conclusion ('We are inclined to think that these limitations are insurmountable in principle, although they might not be'), further showing the conclusion is not forced by definition. No fitted parameters are renamed as predictions, and no uniqueness theorem or ansatz is imported from the authors' prior work. The main weakness is an unproven empirical premise about flat token-string context management, which is a correctness and evidence concern, not a circularity concern.

Assumptions & free parameters 1 free parameters · 4 assumptions · 1 invented entities

The paper introduces no numerical fits and no physical entities. Its dependence is on controversial philosophical theories of context and on unproven empirical claims about LLM architecture. The main free choice is the stipulation of 11 criteria.

free parameters (1)
  • Set and number of CONTEXT-ALIGN criteria = 11 (chosen by hand, not fitted)
    The framework defines conversational alignment as satisfaction of these 11 desiderata, selected by the authors from the philosophical literature. There is no independent derivation that these criteria are necessary or sufficient, so the boundary of the framework is chosen by hand.
assumptions (4)
  • domain assumption Human communication requires tracking context-sensitive content, common ground, a conversational scoreboard, and QUDs as described by Stalnaker, Lewis, Roberts, and Grice.
    Section 2 presents these theories as central despite acknowledging disagreement; the normative force of the framework depends on accepting at least one such theory as a standard for alignment.
  • domain assumption LLMs lack genuine intentionality, situatedness, and rich socio-linguistic competence; their context handling is statistical and mediated by design parameters.
    Section 2.7 asserts this without proof and uses it as the basis for claiming LLMs cannot fully align with human pragmatic norms.
  • domain assumption LLMs manage conversation as a single token string without a structured discourse model, so memory strategies produce either lossy compression or context collapse.
    Section 4.2 uses this premise to derive the fundamental tension between coherence and context collapse.
  • domain assumption Behavioral alignment protocols such as HHH impose a fixed, context-insensitive communicative identity on LLMs.
    Section 5.2 argues that this contrasts with human situated normativity; the claim is supported by examples rather than systematic evidence.
invented entities (1)
  • Static communicative identity of LLMs
    purpose: Explains why behaviorally aligned models cannot adapt self-presentation to context, grounding violations of criteria CA8 and CA2.
    The paper attributes a fixed corporate persona to LLMs (Section 5.2) and claims users cannot reshape it through dialogue. This is asserted with examples, not measured, and no independent falsifiable handle is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Conversational Alignment with Artificial Intelligence in Context." pith.science (2026). https://pith.science/paper/ERYVHBDZ

@misc{pith2026250522907,
  author       = {Pith},
  title        = {Pith review of: Conversational Alignment with Artificial Intelligence in Context},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ERYVHBDZ}},
  note         = {Machine review of arXiv:2505.22907}
}
read the original abstract

The development of sophisticated artificial intelligence (AI) conversational agents based on large language models raises important questions about the relationship between human norms, values, and practices and AI design and performance. This article explores what it means for AI agents to be conversationally aligned to human communicative norms and practices for handling context and common ground and proposes a new framework for evaluating developers' design choices. We begin by drawing on the philosophical and linguistic literature on conversational pragmatics to motivate a set of desiderata, which we call the CONTEXT-ALIGN framework, for conversational alignment with human communicative practices. We then suggest that current large language model (LLM) architectures, constraints, and affordances may impose fundamental limitations on achieving full conversational alignment.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

75 extracted references · 73 canonical work pages

  1. [1]

    The claude 3 model family: Opus, sonnet, haiku

    AI Anthropic. The claude 3 model family: Opus, sonnet, haiku. Claude-3 Model Card, 1: 0 1, 2024

  2. [2]

    Usman Anwar, Abulhair Saparov, Javier Rando, Daniel Paleka, Miles Turpin, Peter Hase, Ekdeep Singh Lubana, Erik Jenner, Stephen Casper, Oliver Sourbut, Benjamin L. Edelman, Zhaowei Zhang, Mario G \"u nther, Anton Korinek, Jose Hernandez-Orallo , Lewis Hammond, Eric Bigelow, Alexander Pan, Lauro Langosco, Tomasz Korbak, Heidi Zhang, Ruiqi Zhong, Se \'a n \...

  3. [3]

    Mind the gap: Challenges of deep learning approaches to Theory of Mind

    Jaan Aru, Aqeel Labash, Oriol Corcoll, and Raul Vicente. Mind the gap: Challenges of deep learning approaches to Theory of Mind . Artificial Intelligence Review, 56 0 (9): 0 9141--9156, September 2023

  4. [4]

    A General Language Assistant as a Laboratory for Alignment , December 2021

    Amanda Askell, Yuntao Bai, Anna Chen, Dawn Drain, Deep Ganguli, Tom Henighan, Andy Jones, Nicholas Joseph, Ben Mann, Nova DasSarma, Nelson Elhage, Zac Hatfield-Dodds , Danny Hernandez, Jackson Kernion, Kamal Ndousse, Catherine Olsson, Dario Amodei, Tom Brown, Jack Clark, Sam McCandlish, Chris Olah, and Jared Kaplan. A General Language Assistant as a Labor...

  5. [5]

    J. L. Austin. How to Do Things with Words . Clarendon Press, Oxford, 1962

  6. [6]

    Faceted Id / Entity : Managing Representation in a Digital World

    danah boyd . Faceted Id / Entity : Managing Representation in a Digital World . PhD thesis, MIT, Cambridge, MA., 2002

  7. [7]

    Social Network Sites as Networked Publics : Affordances , Dynamics , and Implications

    danah boyd . Social Network Sites as Networked Publics : Affordances , Dynamics , and Implications . In Zizi Papacharissi, editor, A Networked Self , pages 39--58. Routledge, New York, N.Y., 2010

  8. [8]

    How ``context collapse'' was coined: My recollection, December 2013

    danah boyd . How ``context collapse'' was coined: My recollection, December 2013

Show all 75 references
  1. [9]

    It's Complicated : The Social Lives of Networked Teens

    danah boyd . It's Complicated : The Social Lives of Networked Teens . Yale University Press, New Haven, CT., 2014

  2. [10]

    Sparks of Artificial General Intelligence : Early experiments with GPT-4 , 2023

    S \'e bastien Bubeck, Varun Chandrasekaran, Ronen Eldan, Johannes Gehrke, Eric Horvitz, Ece Kamar, Peter Lee, Yin Tat Lee, Yuanzhi Li, Scott Lundberg, Harsha Nori, Hamid Palangi, Marco Tulio Ribeiro, and Yi Zhang. Sparks of Artificial General Intelligence : Early experiments w...

  3. [11]

    Adapting Language Models to Compress Contexts

    Alexis Chevalier, Alexander Wettig, Anirudh Ajith, and Danqi Chen. Adapting Language Models to Compress Contexts . In Houda Bouamor, Juan Pino, and Kalika Bali, editors, Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing , pages 3829--3846, ...

  4. [12]

    McKee, Joel Z

    Allan Dafoe, Edward Hughes, Yoram Bachrach, Tantum Collins, Kevin R. McKee, Joel Z. Leibo, Kate Larson, and Thore Graepel. Open Problems in Cooperative AI , December 2020

  5. [13]

    The Epistemic Dangers of Context Collapse Online

    Karen Frost-Arnold . The Epistemic Dangers of Context Collapse Online . In Jennifer Lackey, editor, Applied Epistemology , page 0. Oxford University Press, May 2021

  6. [14]

    Who Should We Be Online ? A Social Epistemology for the Internet

    Karen Frost-Arnold . Who Should We Be Online ? A Social Epistemology for the Internet . Oxford University Press, New York, N.Y., 2023

  7. [15]

    The Challenge of Value Alignment

    Iason Gabriel and Vafa Ghazavi. The Challenge of Value Alignment . In Carissa V \'e liz, editor, The Oxford Handbook of Digital Ethics , pages 336--355. Oxford University Press, Oxford, 2022

  8. [16]

    Gemini: A Family of Highly Capable Multimodal Models , June 2024

    Team Gemini. Gemini: A Family of Highly Capable Multimodal Models , June 2024

  9. [17]

    Improving alignment of dialogue agents via targeted human judgements, September 2022

    Amelia Glaese, Nat McAleese, Maja Tr e bacz, John Aslanides, Vlad Firoiu, Timo Ewalds, Maribeth Rauh, Laura Weidinger, Martin Chadwick, Phoebe Thacker, Lucy Campbell-Gillingham , Jonathan Uesato, Po-Sen Huang, Ramona Comanescu, Fan Yang, Abigail See, Sumanth Dathathri, Rory Gr...

  10. [18]

    The Presentation of Self in Everyday Life

    Erving Goffman. The Presentation of Self in Everyday Life . Doubleday, Garden City, N.Y., 1959

  11. [19]

    Shifting Sands : An Interest-Relative Theory of Vagueness

    Delia Graff. Shifting Sands : An Interest-Relative Theory of Vagueness . Philosophical Topics, 28 0 (1): 0 45--81, 2000

  12. [20]

    Logic and conversation

    Herbert Paul Grice. Logic and conversation. In Peter Cole and Jerry L. Morgan, editors, Syntax and Semantics , 3: Speech Acts , pages 41--58. Academic Press, New York, NY., 1975

  13. [21]

    Laurence R. Horn. Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In Deborah Schiffrin, editor, Meaning, Form , and Use in Context : Linguistic Applications , pages 11--42. Georgetown University Press, Washington, D.C., 1984

  14. [22]

    Demonstratives

    David Kaplan. Demonstratives. In Joseph Almog, John Perry, and Howard Wettstein, editors, Themes from Kaplan , pages 481--563. Oxford University Press, Oxford, 1989

  15. [23]

    In Conversation with Artificial Intelligence : Aligning language Models with Human Values

    Atoosa Kasirzadeh and Iason Gabriel. In Conversation with Artificial Intelligence : Aligning language Models with Human Values . Philosophy & Technology, 36 0 (2): 0 27, April 2023

  16. [24]

    Vagueness and grammar: The semantics of relative and absolute gradable adjectives

    Christopher Kennedy. Vagueness and grammar: The semantics of relative and absolute gradable adjectives. Linguistics and Philosophy, 30 0 (1): 0 45, 2007

  17. [25]

    Alignment of Language Agents , March 2021

    Zachary Kenton, Tom Everitt, Laura Weidinger, Iason Gabriel, Vladimir Mikulik, and Geoffrey Irving. Alignment of Language Agents , March 2021

  18. [26]

    SentimentGPT : Exploiting GPT for Advanced Sentiment Analysis and its Departure from Current Machine Learning , July 2023

    Kiana Kheiri and Hamid Karimi. SentimentGPT : Exploiting GPT for Advanced Sentiment Analysis and its Departure from Current Machine Learning , July 2023

  19. [27]

    Hannah Rose Kirk, Iason Gabriel, Chris Summerfield, Bertie Vidgen, and Scott A. Hale. Why human- AI relationships need socioaffective alignment, February 2025

  20. [28]

    What `must' and `can' must and can mean

    Angelika Kratzer. What `must' and `can' must and can mean. Linguistics and Philosophy, 1 0 (3): 0 337--355, 1977

  21. [29]

    Imagination and Convention : Distinguishing Grammar and Inference in Language

    Ernie Lepore and Matthew Stone. Imagination and Convention : Distinguishing Grammar and Inference in Language . Oxford University Press, Oxford, 2015

  22. [30]

    Presumptive Meanings : The Theory of Generalized Conversational Implicature

    Stephen Levinson, C. Presumptive Meanings : The Theory of Generalized Conversational Implicature . MIT Press, Cambridge, MA., 2000

  23. [31]

    Convention: A Philosophical Study

    David Lewis. Convention: A Philosophical Study . Harvard University Press, Cambridge, MA., 1969

  24. [32]

    General semantics

    David Lewis. General semantics. Synthese, 22: 0 18--67, 1970

  25. [33]

    Counterfactuals

    David Lewis. Counterfactuals. Harvard University Press, Cambridge, MA, 1973

  26. [34]

    Scorekeeping in a language game

    David Lewis. Scorekeeping in a language game. Journal of Philosophical Logic, 8 0 (1): 0 339--359, 1979

  27. [35]

    Karen S. Lewis. Imagined Audiences and Common Ground . In Patrick Connolly, Sanford C. Goldberg, and Jennifer Saul, editors, Conversations Online . Oxford University Press, Oxford, 2025

  28. [36]

    u ttler, Mike Lewis, Wen-tau Yih, Tim Rockt \

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich K \"u ttler, Mike Lewis, Wen-tau Yih, Tim Rockt \"a schel, Sebastian Riedel, and Douwe Kiela. Retrieval- Augmented Generation for Knowledge-Intensive NLP Tasks , April 2021

  29. [37]

    Solving Quantitative Reasoning Problems with Language Models , June 2022

    Aitor Lewkowycz, Anders Andreassen, David Dohan, Ethan Dyer, Henryk Michalewski, Vinay Ramasesh, Ambrose Slone, Cem Anil, Imanol Schlag, Theo Gutman-Solo , Yuhuai Wu, Behnam Neyshabur, Guy Gur-Ari , and Vedant Misra. Solving Quantitative Reasoning Problems with Language Models...

  30. [38]

    Pragmatics in the Era of Large Language Models : A Survey on Datasets , Evaluation , Opportunities and Challenges , February 2025

    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 , February 2025

  31. [39]

    The Code That Binds Us : Navigating the Appropriateness of Human-AI Assistant Relationships

    Arianna Manzini, Geoff Keeling, Lize Alberts, Shannon Vallor, Meredith Ringel Morris, and Iason Gabriel. The Code That Binds Us : Navigating the Appropriateness of Human-AI Assistant Relationships . Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 7 0 (1): 0 ...

  32. [40]

    Marwick and danah boyd

    Alice E. Marwick and danah boyd . I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13 0 (1): 0 114--133, February 2011

  33. [41]

    Context Collapse Online

    Lucy McDonald. Context Collapse Online . In Patrick Connolly, Sanford C. Goldberg, and Jennifer Saul, editors, Conversations Online . Oxford University Press, Oxford, 2025

  34. [42]

    No Sense of Place : The Impact of Electronic Media on Social Behavior

    Joshua Meyrowitz. No Sense of Place : The Impact of Electronic Media on Social Behavior . Oxford University Press, New York, N.Y., 1985

  35. [43]

    A Philosophical Introduction to Language Models -- Part I : Continuity With Classic Debates , January 2024

    Rapha \"e l Milli \`e re and Cameron Buckner. A Philosophical Introduction to Language Models -- Part I : Continuity With Classic Debates , January 2024

  36. [44]

    Pragmatics

    Richard Montague. Pragmatics. In R Klibansky, editor, Contemporary Philosophy . A Study , pages 102--122. La Nuova Italia Editrice, Florence, 1968

  37. [45]

    English as a Formal Language

    Richard Montague. English as a Formal Language . In Bruno Visentini, editor, Linguaggi Nella Societa e Nella Tecnica , pages 188--221. Edizioni di Communita, 1970

  38. [46]

    GPT-4 Technical Report , 2023

    OpenAI. GPT-4 Technical Report , 2023

  39. [47]

    The pragmatics of what is said

    Fran c ois Recanati. The pragmatics of what is said. Mind & Language, 4 0 (4): 0 295--329, 1989

  40. [48]

    Information structure in discourse: Towards an integrated formal theory of pragmatics

    Craige Roberts. Information structure in discourse: Towards an integrated formal theory of pragmatics. Semantics & Pragmatics, 5: 0 1--69, 2012

  41. [49]

    Neural Theory-of-Mind ? On the Limits of Social Intelligence in Large LMs , April 2023

    Maarten Sap, Ronan LeBras, Daniel Fried, and Yejin Choi. Neural Theory-of-Mind ? On the Limits of Social Intelligence in Large LMs , April 2023

  42. [50]

    Thrilled by Your Progress ! Large Language Models ( GPT-4 ) No Longer Struggle to Pass Assessments in Higher Education Programming Courses

    Jaromir Savelka, Arav Agarwal, Marshall An, Chris Bogart, and Majd Sakr. Thrilled by Your Progress ! Large Language Models ( GPT-4 ) No Longer Struggle to Pass Assessments in Higher Education Programming Courses . In Proceedings of the 2023 ACM Conference on International Comp...

  43. [51]

    Schoubye and Andreas Stokke

    Anders J. Schoubye and Andreas Stokke. What is Said ? No \^u s , 50 0 (4): 0 759--793, 2016

  44. [52]

    John R. Searle. Speech Acts . Cambridge University Press, Cambridge, 1969

  45. [53]

    John R. Searle. The Background of Meaning . In John R. Searle, Ferenc Kiefer, and Manfred Bierwisch, editors, Speech Act Theory and Pragmatics , pages 221--232. Springer Netherlands, Dordrecht, 1980

  46. [54]

    John R. Searle. Intentionality: An Essay in the Philosophy of Mind . Cambridge University Press, Cambridge, 1983

  47. [55]

    Learning by Distilling Context , September 2022

    Charlie Snell, Dan Klein, and Ruiqi Zhong. Learning by Distilling Context , September 2022

  48. [56]

    Relevance: Communication and Cognition

    Dan Sperber and Deirdre Wilson. Relevance: Communication and Cognition . Blackwell, Oxford, 1986

  49. [57]

    PUB : A Pragmatics Understanding Benchmark for Assessing LLMs ' Pragmatics Capabilities

    Settaluri Sravanthi, Meet Doshi, Pavan Tankala, Rudra Murthy, Raj Dabre, and Pushpak Bhattacharyya. PUB : A Pragmatics Understanding Benchmark for Assessing LLMs ' Pragmatics Capabilities . In Lun-Wei Ku, Andre Martins, and Vivek Srikumar, editors, Findings of the Association ...

  50. [58]

    A theory of conditionals

    Robert Stalnaker. A theory of conditionals. In N Rescher, editor, Studies in Logical Theory , pages 98--112. Oxford University Press, Oxford, 1968

  51. [59]

    Assertion

    Robert Stalnaker. Assertion. Syntax and Semantics, 9: 0 315--332, 1978

  52. [60]

    Context and Content : Essays on Intentionality in Speech and Thought

    Robert Stalnaker. Context and Content : Essays on Intentionality in Speech and Thought . Oxford University Press, Oxford, 1999

  53. [61]

    Common ground

    Robert Stalnaker. Common ground. Linguistics and Philosophy, 25: 0 701--721, 2002

  54. [62]

    Robert Stalnaker. Context. Oxford University Press, Oxford, 2014

  55. [63]

    On quantifier domain restriction

    Jason Stanley and Zolt \'a n Gendler Szab \'o . On quantifier domain restriction. Mind & Language, 15 0 (2 and 3): 0 219--261, 2000

  56. [64]

    Generics in context

    Rachel Katharine Sterken. Generics in context. Philosophers' Imprint, 15 0 (21): 0 1--30, 2015

  57. [65]

    LLM Theory of Mind and Alignment : Opportunities and Risks , May 2024

    Winnie Street. LLM Theory of Mind and Alignment : Opportunities and Risks , May 2024

  58. [66]

    Winnie Street, John Oliver Siy, Geoff Keeling, Adrien Baranes, Benjamin Barnett, Michael McKibben, Tatenda Kanyere, Alison Lentz, Blaise Aguera y Arcas, and Robin I. M. Dunbar. LLMs achieve adult human performance on higher-order theory of mind tasks, May 2024

  59. [67]

    Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks , March 2023

    Tomer Ullman. Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks , March 2023

  60. [68]

    Attention is All you Need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, ukasz Kaiser, and Illia Polosukhin. Attention is All you Need . Advances in Neural Information Processing Systems, pages 5998--6008, 2017

  61. [69]

    Document- Level Machine Translation with Large Language Models , October 2023

    Longyue Wang, Chenyang Lyu, Tianbo Ji, Zhirui Zhang, Dian Yu, Shuming Shi, and Zhaopeng Tu. Document- Level Machine Translation with Large Language Models , October 2023

  62. [70]

    Qiaosi Wang and Ashok K. Goel. Mutual Theory of Mind for Human-AI Communication , May 2024

  63. [71]

    In- Context Former : Lightning-fast Compressing Context for Large Language Model , November 2024

    Xiangfeng Wang, Zaiyi Chen, Zheyong Xie, Tong Xu, Yongyi He, and Enhong Chen. In- Context Former : Lightning-fast Compressing Context for Large Language Model , November 2024

  64. [72]

    Taxonomy of Risks posed by Language Models

    Laura Weidinger, Jonathan Uesato, Maribeth Rauh, Conor Griffin, Po-Sen Huang, John Mellor, Amelia Glaese, Myra Cheng, Borja Balle, Atoosa Kasirzadeh, Courtney Biles, Sasha Brown, Zac Kenton, Will Hawkins, Tom Stepleton, Abeba Birhane, Lisa Anne Hendricks, Laura Rimell, William...

  65. [73]

    Fiore, and Florian Jentsch

    Jessica Williams, Stephen M. Fiore, and Florian Jentsch. Supporting Artificial Social Intelligence With Theory of Mind . Frontiers in Artificial Intelligence, 5, February 2022

  66. [74]

    Prompt Compression and Contrastive Conditioning for Controllability and Toxicity Reduction in Language Models

    David Wingate, Mohammad Shoeybi, and Taylor Sorensen. Prompt Compression and Contrastive Conditioning for Controllability and Toxicity Reduction in Language Models . In Yoav Goldberg, Zornitsa Kozareva, and Yue Zhang, editors, Findings of the Association for Computational Ling...

  67. [75]

    Philosophical Investigations

    Ludwig Wittgenstein. Philosophical Investigations . Wiley-Blackwell, Oxford, 1953

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.