Pith. sign in

REVIEW 4 major objections 6 minor 64 references

Dynamic Epistemic Friction in Dialogue

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that dynamic epistemic friction, quantified as vector misalignment, predicts how participants revise beliefs in a collaborative dialogue to within a few grams.

desk verdict A promising conceptual framework, undermined by an update rule that contradicts its own worked examples and an evaluation that fits rather than predicts. read the letter →

arxiv 2506.10934 v1 pith:5W2GJGXL submitted 2025-06-12 cs.CL

Timothy Obiso , Kenneth Lai , Abhijnan Nath , Nikhil Krishnaswamy , James Pustejovsky This is my paper
classification cs.CL
keywords dynamicepistemicfrictionbeliefrevisionlogicvectorsymbolicarchitecturesholographicreducedrepresentationscommongrounddialoguestatetrackingWeightsTaskDataset
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 tries to establish that dynamic epistemic friction—the resistance a person feels when new information conflicts with, or sits far from, what they already believe—is a measurable quantity that can be used to predict how beliefs change in dialogue. The authors formalize friction inside Dynamic Epistemic Logic as persistent misalignment between a belief state and an incoming proposition, then encode belief states and assertions as five-dimensional vectors so that alignment becomes a geometric similarity. On four annotated sessions of a collaborative block-weighing task, they feed a participant's belief vector through an update rule with two tuned friction coefficients as each interlocutor utterance arrives; after a fitted linear map, the final vector predicts the true block weights to within 2-3 grams RMSE for three of the four groups. The intended consequence is that resistance to belief revision is not an obstacle to be smoothed away but a signal that dialogue systems can measure, tune, and exploit.

What carries the argument

The engine is a vector-space operationalization of Dynamic Epistemic Logic. Belief states and propositions live in $\mathbb{R}^5$ with components ordered [red, blue, green, purple, yellow]; an assertion such as red=10 and blue=10 is written $[10,10,0,0,0]$, and a denial of a claimed weight contributes a negative entry. The load-bearing update rule is $$\vec{\phi}'_a = \vec{\phi}_a + \min\{\$\beta$,\$\alpha$\,s\}\cdot \mathrm{CosSim}(\vec{\phi}_a,\vec{\phi}_b)\odot \vec{\phi}_b,\qquad s=\mathrm{sgn}(\vec{\phi}_a\cdot \vec{\phi}_b),$$ where $\odot$ is elementwise multiplication, $\alpha$ is the strength of the friction force, and $\beta$ caps how much an aligned assertion can reinforce an existing belief. Friction itself is defined as one minus the cosine-similarity alignment between the belief vector and the proposition-plus-evidence vector, and the QBank-to-EBank-to-FBank structure from evidence-based DEL supplies the vocabulary of where propositions sit as evidence accumulates. The coefficients $\alpha$ and $\beta$ are the degrees of freedom that make the model resist or accelerate belief change, and the paper grid-searches them, finding $\alpha=5$, $\beta=2$ best on the Weights Task data.

What would settle it

Execute equation (17) as written on the paper's own second worked example: with $\vec{\phi}_a=[10,10,20,0,0]$, $\vec{\phi}_b=[10,-10,20,0,0]$, and $\alpha=\beta=1$, the formula scales every nonzero component of $\vec{\phi}_b$ by the same factor and adds it to $\vec{\phi}_a$, so the red and green components should move away from their starting values. The paper reports the updated vector as $[10,3.333,20,0,0]$, leaving red and green exactly unchanged. Recomputing this one line settles whether the published update rule is the rule that produced the reported predictions.

Watch

Extended reading notes

Core claim

The paper's central claim is that epistemic friction can be turned from a metaphor into a computational quantity: define friction as one minus the alignment between an agent's belief vector and the incoming proposition plus its evidence, then update the belief vector by a step whose size and sign depend on that alignment. Concretely, the model predicts the final belief state of the least-talkative participant in each group of the Weights Task by starting from a uniform initial vector, applying the update rule to every interlocutor assertion, and regressing the resulting vector onto the known true weights $[10,10,20,30,50]$. The reported root-mean-squared error of 2-3 grams for most groups is offered as evidence that the resistance profile the model computes tracks the way human participants actually revise their beliefs, and that the two friction coefficients control a trade-off between naive acceptance and rigid refusal of new information.

Load-bearing premise

The whole empirical result rests on the assumption that writing beliefs as five hand-assigned numbers and updating them with the cosine-scaled rule (17) is a faithful implementation of the logical belief revision that DEL describes; if the vectorization and the rule drift from the logic, the error bars measure the linear regressor, not epistemic friction.

Editorial extensions

If this is right

  • Belief revision in dialogue becomes tunable: a system can decide how much resistance to apply to each incoming assertion rather than treating every accepted utterance as a flat update.
  • The optimal coefficients show a 'just enough friction' regime: too little resistance lets beliefs chase incorrect intermediate positions, too much makes them rigid, so the two coefficients act as regularizers on prediction error.
  • The same update machinery extends beyond collaborative tasks to adversarial ones such as Diplomacy, where a spike in friction between aligned agents could flag the moment one agent starts deceiving another.
  • The distance between a participant's current belief vector and a proposed update can feed a classifier that predicts the participant's next move—acceptance, counterargument, or a clarifying question—using the gap the paper equates with friction.

Reading between the lines

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

  • Editorial extension: because the final belief vector is passed through a fitted ridge regressor before scoring, the experiment measures whether DEF dynamics correlate with the final belief state, not whether the update rule alone produces gram-accurate weights; a control that feeds only utterance counts per block into the same regressor would show how much of the fit is friction versus mere dialog
  • Editorial extension: the model assumes utterance order matters, so shuffling the utterances in a group's dialogue should change the predicted final state; if predictions are order-insensitive, the 'dynamic' part of DEF is doing little work beyond counting evidence per block.
  • Editorial extension: substituting learned LLM embeddings for the hand-built [red, blue, green, purple, yellow] vectors would test the paper's admitted isotropy caveat; if the update rule fails with real embeddings, the friction concept may still hold but the hand-built representation, not the concept, is what carried the empirical result.
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

4 major / 6 minor

Summary. The paper defines dynamic epistemic friction (DEF) as resistance to belief revision in dialogue, formalizes it in Dynamic Epistemic Logic (DEL) terms and as vector-symbolic alignment, and evaluates a vector update rule on the Weights Task Dataset. Using hand-built R5 vectors for propositions and a ridge regressor from final belief vectors to ground-truth weights, it reports 2–3 g RMSE for most of four groups and concludes that DEF effectively predicts participant belief updates.

Significance. The conceptual move—quantifying belief-revision resistance via vector alignment and relating it to DEL—is interesting and could be useful for friction-aware dialogue systems. The paper is transparent about several limitations (isotropy assumptions, lack of direct competitors) and uses a public dataset, which are strengths. However, the empirical support is currently weak: the update rule is contradicted by its own worked examples, and the reported error is obtained after supervised regression and grid-searched coefficients rather than by a parameter-free DEF prediction. If the rule were corrected and proper baselines added, the framework could be a worthwhile contribution; in its present form the central claim is not reproducible from the manuscript.

major comments (4)
  1. [§5, Eq. (17)] Equation (17) is inconsistent with the worked examples in the same section. For φ_a=[10,10,20,0,0], φ_b=[10,-10,20,0,0], α=β=1, the equation gives φ'_a=[16.667,3.333,33.333,0,0], because CosSim=2/3 and s=+1, so both red and green components are reinforced. The paper instead prints [10,3.333,20,0,0]. The third example similarly gives [10,4.523,30.954,0,0] rather than [10,4.523,20,0,0]. If the experiments in §5.1 use Eq. (17), the implementation is different from the published rule; if they use the rule behind the printed examples, that rule is not specified. Either way, Table 2 cannot be reproduced or interpreted from the paper alone, and no code is released to resolve the ambiguity.
  2. [§5.1, step 4] The evaluation is supervised in a way that partly manufactures the result. The final extracted belief vector is run through a ridge regressor fit to the ground-truth final FBank, and α/β are chosen by grid search (with values 0.01..100 and k up to 4). The reported 2–3 g RMSE therefore measures the joint DEF+ridge+tuned-coefficient pipeline, not the predictive power of DEF itself. A ridge with five features trained on three dialogues can absorb systematic vectorization error; without a no-regression baseline (e.g., raw final vector vs. ground truth, or a constant predictor), the RMSE does not support the claim in §6 that DEF itself effectively predicts belief updates.
  3. [§5.2, Table 2] The empirical basis is four dialogues, and no frictionless or naive baseline is implemented. The text claims a frictionless setting 'would involve a greater error than all models involving friction', but no such baseline appears in Table 2 or Fig. 2; low α/β values are not a frictionless condition because the update still uses Eq. (17) with nonzero coefficients. The conclusion that DEF outperforms frictionless updating is therefore not demonstrated. Additionally, group 4 has RMSE 7.7–17.5 g, so the 'most groups' claim in §5.2 should be qualified.
  4. [§3–§5] The claimed connection to DEL is not operationalized. Equation (17) is a heuristic vector update; the paper does not show that it corresponds to a DEL product update or to the evidence-bank transitions in (7), and axioms (5) are used only for direct component assignment. The R5 vectorization and the update rule are the actual mechanism being credited for prediction, but their relation to the formal DEL model is asserted rather than derived. If the contribution is the DEL-based theory, this gap needs to be filled; if it is the vector heuristic, the framing should be scaled back accordingly.
minor comments (6)
  1. [§6, Fig. 2, §5] There are numerous typos: 'isotopic' should be 'isotropic', 'alledges' should be 'allows', 'peformance' should be 'performance' (also in Fig. 2 captions), and §5 says '4 of of the 10 groups'.
  2. [Eq. (17)] Equation (17) should define the Hadamard product ⊙ and clarify the scope of min; as written, min(β, α×s) could be read as min(β, α)×s.
  3. [Table 2] Table 2 reports intervals like ±0.421 but does not state whether these are standard deviations, standard errors, or confidence intervals, nor over what (iterations, groups, or ridge fits) they are computed.
  4. [§5.1] The procedure says the focus participant is the one who speaks least, but also that 'which specific participant this is may vary across groups and is not further explicated here'; this should be made concrete for reproducibility.
  5. [§5] The representation of inequalities (e.g., yellow<40 as [0,0,0,0,40−U(0,1)]) should be justified; using a random draw in the vectorization introduces stochasticity that is not otherwise acknowledged in the update.
  6. [References] The references are incomplete in places: Markowska et al. lacks year and venue, Ginzburg et al. contains a placeholder '1 others', and several arXiv preprints are cited without version or accession details.

Circularity Check

2 steps flagged · score 6.0 of 10

Final-belief 'prediction' is partly self-definitional: target values enter the feature vector via initialization and Saying-is-Believing direct assignment, and the reported RMSE comes from friction coefficients and a ridge regressor fitted on the same four groups.

  1. self definitional [Sec. 5.1, procedure steps 1-4 and preceding 'Saying is Believing' paragraph]
    "Initialize the focus participant's 'belief vector' from a uniform distribution U(0, 10) and set the first ('red') element of the belief vector to 10. This reflects the initial state of the Weights Task where participants are told that the red block weighs 10g. ... if the focus participant makes a statement asserting a block weight or explicitly accepts another participant's positive and specific assertion about a block weight ..., then that value gets directly assigned to the relevant component in the focus participant's belief vector ..."

    The target vector [10,10,20,30,50] is partly an input: red=10 is set by initialization, and any weight the focus participant states or explicitly accepts is written directly into the belief vector before the ridge regressor is fitted. Thus the regressor maps a feature vector that already contains target values to that same target; for those components, the reported RMSE measures copying input to output, not predicting an unseen belief update. The Sec. 6 conclusion that DEF 'effectively predicts participant belief updates' is therefore, for those components, true by construction rather than by empirical test.

  2. fitted input called prediction [Sec. 5.2, grid-search paragraph and Table 2]
    "We performed a grid search through different values∈{ 0.01..100} with a dialogue history window size k of up to 4, using leave-one-group-out cross-validation. From this search, α =5 and β =2 emerged as the best-performing combination. These values were used to compute the groupwise results in Table 2."

    The friction coefficients α and β are fitted to minimize error on the same four groups whose RMSE is then reported as evidence of DEF's predictive power, and the pipeline additionally fits a ridge regressor to the ground-truth FBank. The Table 2 errors are therefore fit-quality numbers for a model selected on the target data, not parameter-free predictions. The paper's claim that 'DEF effectively predicts participant belief updates' is supported only by the performance of these tuned parameters and the fitted regressor, so the prediction is statistically forced rather than independent.

full rationale

The central empirical claim in Sec. 6 is that DEF 'effectively predicts participant belief updates,' operationalized as predicting the final FBank weights. That claim is only partially supported by an independent derivation. First, the feature vector is initialized with red=10 (the target value for that block) and, via the Saying-is-Believing axiom, any weight the focus participant states or explicitly accepts is directly written into the belief vector; since all WTD groups converge to the ground-truth weights, the ridge regressor's input can contain target values, so for those components the 'prediction' is a copy of an input rather than a forecast of an unseen update. Second, the reported 2-3g RMSE is produced by friction coefficients α=5, β=2 selected by grid search on the same four groups, plus a ridge regressor fitted to the ground-truth FBank; the errors therefore partly measure fit quality. There is also a serious internal inconsistency: Eq. (17) applied to the second worked example yields [16.667, 3.333, 33.333, 0, 0], not the printed [10, 3.333, 20, 0, 0]; the third example likewise would change the green component, so the update rule that actually generated Table 2 is unspecified. This is a correctness and reproducibility risk rather than a circularity per se. Self-citations (Khebour et al. 2024a,b; Zhu et al. 2024) are used for vectorization conventions and evidence-based DEL but are not load-bearing in the sense of importing an unverified uniqueness theorem, so they do not raise the score. Overall, the derivation is not equivalent to its inputs as a whole, but several components of the reported 'prediction' reduce by construction, giving partial circularity.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central evaluation depends on two tuned friction coefficients, a fitted ridge regressor, a hand-built R5 encoding, and a direct-assignment rule for accepted assertions. The DEL machinery is used as conceptual motivation rather than as a derivation of the update rule, so the empirical output is partly fit, not purely predicted.

free parameters (3)
  • Friction force coefficient alpha = 5
    Grid-searched over 0.01 to 100 in Sec. 5.2; scales how strongly an utterance moves the belief vector in Eq. (17).
  • Friction ceiling coefficient beta = 2
    Grid-searched together with alpha; caps reinforcement from aligned utterances in Eq. (17).
  • Ridge regression mapping from final belief vector to ground-truth weights = weights not reported
    Fit on 3 training groups in the leave-one-group-out loop (Sec. 5.1); maps the 5-dimensional (or concatenated) vector to [10,10,20,30,50].
assumptions (4)
  • ad hoc to paper The R5 vectorization of propositions (block weights as coordinate values, zeros as no information) supports arithmetic operations that mirror logical connectives.
    Sec. 5; the authors concede in Sec. 6 that this requires near-isotropy, which is inconsistent in LLMs, and the mapping is task-specific.
  • domain assumption Saying is Believing (Eq. 5b): assertions and explicit acceptances by the focus participant directly set belief vector components.
    Sec. 5.1 relies on direct assignment for positive assertions; this is plausible but not independently tested.
  • domain assumption DEL product updates (Eq. 2) are an adequate formal model for the conversational belief changes studied.
    Sec. 3 uses M tensor E to define updates; the paper gives no empirical check that DEL updates match human revision trajectories beyond the fitted vector update.
  • ad hoc to paper The vector update rule (17) is a valid operationalization of DEL belief revision and its gradient-style step.
    Eq. 17 is posited rather than derived; worked examples do not match the formula as written (red and green components should change).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dynamic Epistemic Friction in Dialogue." pith.science (2026). https://pith.science/paper/5W2GJGXL

@misc{pith2026250610934,
  author       = {Pith},
  title        = {Pith review of: Dynamic Epistemic Friction in Dialogue},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5W2GJGXL}},
  note         = {Machine review of arXiv:2506.10934}
}
read the original abstract

Recent developments in aligning Large Language Models (LLMs) with human preferences have significantly enhanced their utility in human-AI collaborative scenarios. However, such approaches often neglect the critical role of "epistemic friction," or the inherent resistance encountered when updating beliefs in response to new, conflicting, or ambiguous information. In this paper, we define dynamic epistemic friction as the resistance to epistemic integration, characterized by the misalignment between an agent's current belief state and new propositions supported by external evidence. We position this within the framework of Dynamic Epistemic Logic (Van Benthem and Pacuit, 2011), where friction emerges as nontrivial belief-revision during the interaction. We then present analyses from a situated collaborative task that demonstrate how this model of epistemic friction can effectively predict belief updates in dialogues, and we subsequently discuss how the model of belief alignment as a measure of epistemic resistance or friction can naturally be made more sophisticated to accommodate the complexities of real-world dialogue scenarios.

Figures

Figures reproduced from arXiv: 2506.10934 by the authors.

Figure 1
Figure 1. A group of 3 performing the Weights Task. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Effects of different α and β values in the vector update function (17) on DEF peformance in FBank prediction. Values shown are averaged over leave-one-group-out cross-validation. Error bars represent standard error over 100 iterations, after cross-validation. update , error drops dramatically. In other words, without enough friction, beliefs shift too rapidly toward ultimately incorrect positions. With too much, the… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

64 extracted references · 55 canonical work pages

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Nicholas Asher and Anthony Gillies. 2003. Common ground, corrections, and coordination. Argumentation, 17(4):481--512

    work page 2003

  4. [4]

    Marco Baroni. 2013. Composition in distributional semantics. Language and Linguistics Compass, 7(10):511--522

    work page 2013

  5. [5]

    Ringo Baumann and Gerhard Brewka. 2015. Agm meets abstract argumentation: expansion and revision for dung frameworks. In IJCAI, volume 15, pages 2734--2740

    work page 2015

  6. [6]

    Thomas Bolander. 2014. Seeing is believing: Formalising false-belief tasks in dynamic epistemic logic. In European conference on social intelligence (ECSI 2014), pages 87--107

    work page 2014

  7. [7]

    Gemma Boleda. 2020. Distributional semantics and linguistic theory. Annual Review of Linguistics, 6(1):213--234

    work page 2020

  8. [8]

    Brown, Rebecca Hutchinson, Paul N

    Ralf D. Brown, Rebecca Hutchinson, Paul N. Bennett, Jaime G. Carbonell, and Peter Jansen. 2003. https://aclanthology.org/2003.mtsummit-papers.4/ Reducing boundary friction using translation-fragment overlap . In Proceedings of Machine Translation Summit IX: Papers, New Orleans, USA

    work page 2003

Show all 64 references
  1. [9]

    positive friction

    Zeya Chen and Ruth Schmidt. 2024. Exploring a behavioral model of “positive friction” in human-ai interaction. In International Conference on Human-Computer Interaction, pages 3--22. Springer

    2024

  2. [10]

    Herbert H Clark and Deanna Wilkes-Gibbs. 1986. Referring as a collaborative process. Cognition, 22(1):1--39

    1986

  3. [11]

    Kawin Ethayarajh. 2019. How contextual are contextualized word representations? comparing the geometry of bert, elmo, and gpt-2 embeddings. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on N...

    2019

  4. [12]

    Kerstin Fischer. 2011. How people talk with robots: Designing dialog to reduce user uncertainty. AI Magazine, 32(4):31--38

    2011

  5. [13]

    Jonathan Ginzburg, Chris Eliasmith, and Andy L\" u cking. 2024. https://www.semdial.org/anthology/Z24-Ginzburg_semdial_0007.pdf Swann’s name: Towards a dialogical brain semantics . In Proceedings of the SemDial 2024 Workshop, Trento, Italy. ACL. Accessed: 2025-03-14

    2024

  6. [14]

    Jonathan Ginzburg and 1 others. 1996. Dynamics and the semantics of dialogue. Logic, language and computation, 1:221--237

    1996

  7. [15]

    Howard Goldowsky and Vasanth Sarathy. 2024. Analogical reasoning within a conceptual hyperspace. arXiv preprint arXiv:2411.08684

    2024 arXiv

  8. [16]

    Herbert P Grice. 1975. Logic and conversation. In Speech acts, pages 41--58. Brill

    1975

  9. [17]

    Hadley, Graham Naylor, and Antonia F

    Lauren V. Hadley, Graham Naylor, and Antonia F. de C. Hamilton. 2022. https://doi.org/10.1038/s44159-021-00008-w A review of theories and methods in the science of face-to-face social interaction . Nature Reviews Psychology, 1(1):42--54. Number: 1 Publisher: Nature Publishing Group

    2022 doi

  10. [18]

    Anthony Hunter, Sylwia Polberg, and Matthias Thimm. 2020. Epistemic graphs for representing and reasoning with positive and negative influences of arguments. Artificial Intelligence, 281:103236

    2020

  11. [19]

    Julie Hunter, Nicholas Asher, and Alex Lascarides. 2018. A formal semantics for situated conversation. Semantics and Pragmatics, 11

    2018

  12. [20]

    o khan T \

    Mert \.I nan, Anthony Sicilia, Suvodip Dey, Vardhan Dongre, Tejas Srinivasan, Jesse Thomason, G \"o khan T \"u r, Dilek Hakkani-T \"u r, and Malihe Alikhani. 2025. Better slow than sorry: Introducing positive friction for reliable dialogue systems. arXiv preprint arXiv:2501.17348

    2025 arXiv

  13. [21]

    Pentti Kanerva. 1988. Sparse distributed memory. MIT press

    1988

  14. [22]

    Ibrahim Khebour, Richard Brutti, Indrani Dey, Rachel Dickler, Kelsey Sikes, Kenneth Lai, Mariah Bradford, Brittany Cates, Paige Hansen, Changsoo Jung, Brett Wisniewski, Corbyn Terpstra, Leanne M Hirshfield, Sadhana Puntambekar, Nathaniel Blanchard, James Pustejovsky, and Nikhi...

    2024

  15. [23]

    Ibrahim Khebour, Kenneth Lai, Mariah Bradford, Yifan Zhu, Richard Brutti, Christopher Tam, Jingxuan Tu, Benjamin Ibarra, Nathaniel Blanchard, Nikhil Krishnaswamy, and James Pustejovsky. 2024 b . Common ground tracking in multimodal dialogue. arXiv preprint arXiv:2403.17284

    2024 arXiv

  16. [24]

    Denis Kleyko, Dmitri Rachkovskij, Evgeny Osipov, and Abbas Rahimi. 2023. https://doi.org/10.1145/3558000 A survey on hyperdimensional computing aka vector symbolic architectures, part ii: Applications, cognitive models, and challenges . ACM Computing Surveys, 55(9):1–52

    2023 doi

  17. [25]

    Rachkovskij, Evgeny Osipov, and Abbas Rahimi

    Denis Kleyko, Dmitri A. Rachkovskij, Evgeny Osipov, and Abbas Rahimi. 2022. https://doi.org/10.1145/3538531 A survey on hyperdimensional computing aka vector symbolic architectures, part i: Models and data transformations . ACM Computing Surveys, 55(6):1–40

    2022 doi

  18. [26]

    Nikhil Krishnaswamy and James Pustejovsky. 2020. A formal analysis of multimodal referring strategies under common ground. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 5919--5927

    2020

  19. [27]

    Geert-Jan M Kruijff, Pierre Lison, Trevor Benjamin, Henrik Jacobsson, Hendrik Zender, Ivana Kruijff-Korbayov \'a , and Nick Hawes. 2010. Situated dialogue processing for human-robot interaction. In Cognitive systems, pages 311--364. Springer

    2010

  20. [28]

    Alessandro Lenci and Magnus Sahlgren. 2023. Distributional semantics. Cambridge University Press

    2023

  21. [29]

    Jiaming Luo, Yuan Cao, and Yonghui Wu. 2018. Towards decomposed linguistic representation with holographic reduced representation

    2018

  22. [30]

    Anemily Machina and Robert Mercer. 2024. Anisotropy is not inherent to transformers. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 4892--4907

    2024

  23. [31]

    Formal representation of common ground in dialogue

    Magdalena Markowska, Adil Soubki, Gary Mar, Seyed Abolghasem Mirroshandel, Owen Rambow, and Anita Wasilewska. Formal representation of common ground in dialogue

  24. [32]

    Abhijnan Nath, Carine Graff, Andrei Bachinin, and Nikhil Krishnaswamy. 2025. Frictional A gent A lignment F ramework: S low D own and D on’t B reak T hings. In Annual Meeting of the Association for Computational Linguistics (ACL). ACL

    2025

  25. [33]

    Abhijnan Nath, Sheikh Mannan, and Nikhil Krishnaswamy. 2023. Axomiyaberta: A phonologically-aware transformer model for assamese. In Findings of the Association for Computational Linguistics: ACL 2023, pages 11629--11646

    2023

  26. [34]

    Vlad Niculae, Srijan Kumar, Jordan Boyd-Graber, and Cristian Danescu-Niculescu-Mizil. 2015. Linguistic harbingers of betrayal: A case study on an online strategy game. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th Intern...

    2015

  27. [35]

    Timothy Obiso. 2024. Holographic embeddings for text and graphs a master’s thesis presented to the faculty of the graduate school of arts and sciences

    2024

  28. [36]

    Xenia Ohmer, Marko Duda, and Elia Bruni. 2022. Emergence of hierarchical reference systems in multi-agent communication. In Proceedings of the 29th International Conference on Computational Linguistics, pages 5689--5706

    2022

  29. [37]

    Derek Palmer, Yifan Zhu, Kenneth Lai, Hannah VanderHoeven, Mariah Bradford, Ibrahim Khebour, Carlos Mabrey, Jack Fitzgerald, Nikhil Krishnaswamy, Martha Palmer, and 1 others. 2024. Speech is not enough: Interpreting nonverbal indicators of common knowledge and engagement. arXi...

    2024 arXiv

  30. [38]

    Tony A Plate. 1995. Holographic reduced representations. IEEE Transactions on Neural networks, 6(3):623--641

    1995

  31. [39]

    James Pustejovsky and Nikhil Krishnaswamy. 2021. Embodied human computer interaction. KI-K \"u nstliche Intelligenz , 35(3):307--327

    2021

  32. [40]

    James Pustejovsky and Nikhil Krishnaswamy. 2025. Frictive policy optimization for llm agent interactions. In Proc. of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2025)

    2025

  33. [41]

    Matthias Scheutz, Rehj Cantrell, and Paul Schermerhorn. 2011. Toward humanlike task-based dialogue processing for human robot interaction. Ai Magazine, 32(4):77--84

    2011

  34. [42]

    Robert Stalnaker. 2002. Common ground. Linguistics and philosophy, 25(5/6):701--721

    2002

  35. [43]

    David Traum. 1994. A computational theory of grounding in natural language conversation. PhD thesis, University of Rochester

    1994

  36. [44]

    David R Traum and Staffan Larsson. 2003. The information state approach to dialogue management. Current and new directions in discourse and dialogue, pages 325--353

    2003

  37. [45]

    Johan Van Benthem and Eric Pacuit. 2011. Logical dynamics of evidence. In International Workshop on Logic, Rationality and Interaction, pages 1--27. Springer

    2011

  38. [46]

    Johan van Benthem and Sonja Smets. 2015. Dynamic logics of belief change

    2015

  39. [47]

    Hans Van Ditmarsch, Wiebe van Der Hoek, and Barteld Kooi. 2007. Dynamic epistemic logic, volume 337. Springer Science & Business Media

    2007

  40. [48]

    Hannah VanderHoeven, Brady Bhalla, Ibrahim Khebour, Austin Youngren, Videep Venkatesha, Mariah Bradford, Jack Fitzgerald, Carlos Mabrey, Jingxuan Tu, Yifan Zhu, Kenneth Lai, Changsoo Jung, James Pustejovsky, and Nikhil Krishnaswamy. 2025. TRACE: Real-Time Multimodal Common Gro...

    2025

  41. [49]

    Videep Venkatesha, Abhijnan Nath, Ibrahim Khebour, Avyakta Chelle, Mariah Bradford, Jingxuan Tu, James Pustejovsky, Nathaniel Blanchard, and Nikhil Krishnaswamy. 2024. Propositional extraction from natural speech in small group collaborative tasks. In Proceedings of the 17th I...

    2024

  42. [50]

    Jason D Williams, Antoine Raux, and Matthew Henderson. 2016. The dialog state tracking challenge series: A review. Dialogue & Discourse, 7(3):4--33

    2016

  43. [51]

    Wichayaporn Wongkamjan, Feng Gu, Yanze Wang, Ulf Hermjakob, Jonathan May, Brandon M Stewart, Jonathan Kummerfeld, Denis Peskoff, and Jordan Boyd-Graber. 2024. More victories, less cooperation: Assessing cicero’s diplomacy play. In Proceedings of the 62nd Annual Meeting of the ...

    2024

  44. [52]

    Yifan Zhu, Hannah VanderHoeven, Kenneth Lai, Mariah Bradford, Christopher Tam, Ibrahim Khebour, Richard Brutti, Nikhil Krishnaswamy, and James Pustejovsky. 2024. Modeling theory of mind in multimodal hci. In International Conference on Human-Computer Interaction, pages 205--22...

    2024

  45. [53]

    Baltag, L.S

    A. Baltag, L.S. Moss, and S. Solecki. The logic of public announcements, common knowledge, and private suspicions. In Proceedings of TARK , 1998

    1998

  46. [54]

    R. Gayler. Vector symbolic architectures answer Jackendoff’s challenges for cognitive neuroscience. In Joint International Conference on Cognitive Science , 2003

    2003

  47. [55]

    H.P. Grice. Logic and conversation. In Syntax and Semantics 3: Speech Acts , pages 41–58, 1975

    1975

  48. [56]

    Hintikka

    J. Hintikka. Knowledge and Belief . Cornell University Press, 1962

    1962

  49. [57]

    Khebour, K

    I. Khebour, K. Lai, M. Bradford, Mariah, Y. Zhu, R. Brutti, C. Tam, J. Tu, B. Ibarra, N. Blanchard, K. Krishnaswamy, J. Pustejovsky. Common Ground Tracking in Multimodal Dialogue In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language R...

    2024

  50. [58]

    McDowell

    J. McDowell. Mind and World . Harvard University Press, 1994

    1994

  51. [59]

    T. Plate. Holographic reduced representations. IEEE Transactions on Neural Networks , 6(3):623–641, 1995

    1995

  52. [60]

    S. Sher. Epistemic friction: Reflections on knowledge, truth, and logic. Philosophy and Phenomenological Research , 69(2):294–315, 2004

    2004

  53. [61]

    Sreedharan, A

    S. Sreedharan, A. Kulkarni, S. Kambhampati. Explainable Human-AI Interaction: A Planning Perspective. Springer Nature, 2022

    2022

  54. [62]

    Stalnaker

    R. Stalnaker. Common ground. Linguistics and Philosophy , 25(5-6):701–721, 2002

    2002

  55. [63]

    Van Benthem, E

    J. Van Benthem, E. Pacuit. Logical dynamics of evidence. In International Workshop on Logic, Rationality and Interaction, pages 1--27. Springer, 2011

    2011

  56. [64]

    Y. Zhu, H. VanderHoeven, K. Lai, M. Bradford, C. Tam, I. Khebour, R. Brutti, N. Krishnaswamy, J. Pustejovsky Modeling Theory of Mind in Multimodal HCI International Conference on Human-Computer Interaction, 2024

    2024

Pith tools

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