Pith. sign in

REVIEW 3 major objections 5 minor 35 references

When Do Institutions Beat Intelligence?

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper claims that institutional rules for collective state construction beat additional agent capability only when they repair a structural failure in routing, evidence admission, checkable enforcement, or action representation, and…

desk verdict A serious, mostly sound empirical boundary on when institutions beat scaling in multi-agent LLMs; the frontier-substitution panel needs arm means before the boldest claim is fully supported. read the letter →

arxiv 2608.11357 v1 pith:UNWSICY6 submitted 2026-08-11 cs.MA

Zhengye Han This is my paper · ORCID
classification cs.MA
keywords multi-agentsystemsinstitutionscollectiveintelligencegroupdecisionmakinglargelanguagemodelagentsempiricalmethodologydistributedcognitioncheckability
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

More capable agents do not automatically form a more capable collective, and this paper tries to pinpoint when the fix should be a stronger reasoner and when it should be a change in the rules through which agents pool information. Using controlled artificial ecologies built around four failure loci—access and routing, admission and dependence, state maintenance and incentives, and representation and action—it finds a consistent boundary: institutions help when they repair how the collective constructs a usable public state. The advantage disappears when the institutional signal is uninformative or uncheckable, when stronger intelligence can perform the same transformation directly, or when the constructed state cannot support reliable action. The upshot is that choosing between intelligence and institutions should be a diagnosis of where collective reasoning fails, not a general preference for one over the other.

What carries the argument

The carrying object is the formal institution as state-construction rule: $I = \langle \rho, V, U, \Phi \rangle$, where $\rho$ is who may observe, report, or route; $V$ is which reports are admitted as public evidence; $U$ is how admitted evidence updates shared state; and $\Phi$ is how that state is represented to the finalizer. The paper pairs this with two controlled comparisons: a mechanism-validity interaction $\Gamma_I(m) = \tau_I^+(m) - \tau_I^-(m)$, which asks whether performance tracks the informational signal the institution claims to use, and a crossover $C_I(m_s, m_\ell)$ that compares an institutionally supported weaker reasoner with a stronger raw reasoner under matched calls. These definitions, applied inside ecologies built around access, admission, maintenance, and representation, turn "institutions versus intelligence" into a test of whether the proposed rule is informative, checkable, substitutable, and executable.

What would settle it

Give the stronger raw reasoner on the distributed-evidence ecology the same token budget, latency, or dollar cost as the institutional pipeline instead of only the same call count, and measure whether the +0.58 paired advantage persists; if the advantage collapses while routing completeness remains unchanged, the claim that institutions beat intelligence by repairing coverage rather than by spending more computation is undercut.

Watch

Extended reading notes

Core claim

The paper's central discovery is that whether institutions beat intelligence is an empirical, locatable fact rather than a stylistic preference. Across four controlled ecologies, an institution—a fixed rule for who may report, what is admitted into public belief, how shared state is updated, and how it is exposed to the final decision—improves collective accuracy only when it repairs a specific failure along the path from private observation to public action. Concretely: validated, coverage-complete routing lets a five-call 8B model beat a five-call raw 70B model by +0.58 paired accuracy on a distributed-evidence task; grounded admission takes faulty-evidence accuracy from 0.00 to 1.00; a 10% checkable audit at capability 0.55 beats an unaudited capability 0.65 by +0.135, while all four sanction levels at zero checkability produce exactly zero gain; and a learned reranker outperforms a hand-built graph, while a fixed-evidence interface reversal shows that better evidence does not guarantee better execution. The same boundary conditions define the limits: uninformative or uncheckable signals, capable raw reasoners that can perform the transformation directly, and unexecutable representations all erase the institutional advantage.

Load-bearing premise

The attribution of gains to institutional repair rather than to extra compute or extra information assumes that matching the number of reasoning calls, together with mechanism-breaking controls, isolates the institutional mechanism; the paper itself notes that call matching is not the same as matching tokens, latency, or monetary cost.

Editorial extensions

If this is right

  • In a workflow whose failure is poor evidence routing, buying a larger or more powerful model is the wrong intervention: the distributed-evidence results show complete routing plus validation converts a 0.26-accuracy 3x2 assignment into 1.00 under the disjoint 4x4 assignment, whereas call-matched raw strength on the same partial state stays near the floor.
  • An institutional rule earns causal credit only when performance tracks its signal: syntax-only validation, broken credentials, and uncheckable sanctions produce no gain, so adding such machinery is pure overhead.
  • Institutional advantage is capability-relative: as models or learned retrieval components become able to perform the same transformation from raw context, the permanent institution becomes unnecessary, as shown by the frontier substitution and learned-reranker-over-graph results.
  • Construction and representation must be evaluated separately: improving the public state's evidence quality does not guarantee better final decisions, as shown by the fixed-evidence interface reversal where the ordered interface flips the graph comparison.
  • The practical rule of thumb the paper supports: buy intelligence when capability can perform the transformation; build an institution when an active structural failure survives scaling, the rule is checkable, and the resulting state is executable; otherwise redesign the interface.

Reading between the lines

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

  • A natural extension the paper does not run: an adaptive controller that measures which failure locus is alive (routing coverage, admission validity, checkability, interface executability) and switches between raw reasoning and institutional machinery on that diagnosis; the paper's Gamma and crossover comparisons are exactly the hooks such a controller would need.
  • Because the paper's resource accounting stops at call counts, a reader should expect institutional advantages to shrink when compared at equal token, latency, or dollar budgets; if they do, the boundary would be economic rather than purely cognitive.
  • The zero-checkability exact-zero result suggests a testable design rule for LLM workflows: any governance mechanism whose outputs cannot be independently checked by downstream agents should be predicted to contribute zero marginal accuracy, which could be probed by varying only the verifiability of a reviewer's output while holding its content distribution fixed.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper asks under what conditions institutional design outperforms additional model capability in multi-agent LLM collectives. It formalizes an institution as externally specified rules for observation, admission, state updating, and action interface, and constructs four synthetic ecologies around access/routing, admission/dependence, state maintenance/incentives, and representation/action. Positive institutional interventions are paired with call-matched raw baselines and mechanism-breaking controls, with additional tests on HotpotQA, MuSiQue, and frontier models. The central claim is that institutions help when they repair failures in constructing usable public state, but lose their advantage when their signals are uninformative or uncheckable, when stronger capability can perform the same transformation directly, or when the resulting state cannot support reliable action. The paper is explicitly framed as a diagnosis of where collective reasoning fails rather than as a universal ranking of workflows.

Significance. If the results hold, the paper offers a valuable empirical methodology for deciding when to invest in model capability versus institutional structure in multi-agent systems. Its strengths include controlled manipulations with paired baselines, mechanism-breaking controls (broken routing, syntax-only validation, zero checkability, fixed-evidence interfaces), honest reporting of nulls and wide intervals, and transfer to natural tasks (HotpotQA, MuSiQue). The manuscript is falsifiable in spirit, and it explicitly does not claim that institutions are universally superior. The taxonomy of collective failure loci, and the insistence on separating construction of public state from execution on it, are useful contributions regardless of the specific point estimates.

major comments (3)
  1. [Section 7, Fig. 8a, Table 1] The zero valid-minus-broken interaction for Claude in all three frontier ecologies is interpreted as evidence that a capable reasoner can reliably perform the same transparent transformation from raw context. With only the interaction reported, this zero is compatible with at least three states: high accuracy in both arms (substitution), low accuracy in both arms (the model ignores the institutional signal), and a floor or ceiling null. The abstract's third boundary condition, that institutions lose their advantage when stronger intelligence can perform the transformation directly, rests on this panel, so the arm means for the valid and broken conditions for each frontier model must be reported, or an accuracy measure for the direct transformation must be provided. If Claude's zero interaction is actually low-and-equal, the result belongs under the 'uninformative signal' boundary rather than a capability-substitution boundary. The MuSiQue learned-reranker comparison supports substitution only for evidence construction, not for the dynamic-state, expertise-routing, or reporting transformations tested in Fig. 8a.
  2. [Section 4, Fig. 4 and below-threshold routing] The 3x2 and 3x4 routing conditions vary the number of clue cards exposed (6 and 12 cards, respectively) alongside routing completeness, so the reported chain 'routing completeness -> decisive-evidence coverage -> unique public decision state -> final accuracy' is confounded with information quantity. The claim that below-threshold routing tests 'whether coverage rather than workflow complexity matters' is not cleanly supported unless card count is held fixed across routing conditions. The paper should state explicitly whether the raw baselines in the headline +0.58 comparison also receive all 16 clue cards under matched calls; if they do not, the institutional gain is not separated from simply showing more evidence to the finalizer.
  3. [Section 2, Eq. (1)] The formal definition of the mechanism-validity interaction is not fully specified. The notation I^r_0 is never defined, and as written tau^r_I(m) is a difference relative to an unspecified baseline, so Gamma_I(m) is not transparently equal to the valid-minus-broken interaction used in the experiments. Because this definition is introduced as the justification for giving mechanistic credit only when performance tracks the institutional signal, please define the baseline institution explicitly or remove the formal equations in favor of the verbal definition used in the experimental sections.
minor comments (5)
  1. [Section 7, fixed-evidence interface] In the sentence 'the same passages are presented either through the graph-style public state or through an ordered or passage-pointer interface,' it is unclear whether the ordered interface and the passage-pointer interface are the same condition or two different conditions; please clarify and state whether finalizer call counts are matched in this fixed-evidence comparison.
  2. [Figure 4 caption] The caption says points are arm accuracies with n=50 and that the bottom row reports the paired valid-institution minus stronger-raw effect; adding the call count to each labeled method row (e.g., 5-call 8B vs 5-call 70B) would make the matched-calls claim easier to verify from the figure alone.
  3. [Section 1 and Section 10, matched calls] The limitation paragraph concedes that call matching does not equal token, latency, monetary cost, or engineering burden; this caveat is important framing and should also appear where 'matched reasoning baselines' is first introduced in Section 1, to avoid leaving the impression that the comparison is a matched-resource comparison.
  4. [Section 3 and Section 7, HotpotQA sample size] HotpotQA uses 20 paired questions per model; given the small sample, the paper should state this sample size explicitly in the HotpotQA subsection and connect it to the reported bootstrap intervals, since the graph-routing answer effects may be wide.
  5. [Section 3, model naming] The names 'Qwen 235B', 'Mistral 24B', and 'Claude Sonnet 4.6' appear without version qualifiers; please give the exact model identifiers used in the API calls so that the results are reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: central claims rest on controlled comparisons and external benchmarks, not on definitional identity or self-citation.

full rationale

The paper's derivation chain is empirical rather than definitional: it formally defines institutions in Section 2 but does not derive any performance prediction from that definition alone. Every central claim is supported by paired controlled manipulations with positive, broken, and matched baselines (valid-minus-broken interactions, call-matched raw reasoners, fixed-evidence controls), and the boundary conditions (uninformative signals, zero checkability, substitution by stronger models or learned retrieval, execution-interface reversals) are results that could have gone the other way and are anchored to external tasks (HotpotQA, MuSiQue). There is no fitted parameter later renamed as a prediction, no self-citation chain (the reference list contains no author self-citations), and no uniqueness theorem imported from prior work by the same author. The only conceptual proximity is that the institution definition and the failure taxonomy share the same functional vocabulary, so the summary that institutions help when they repair state construction is a natural restatement; however, that is a modeling choice, not an equation that forces the outcome. The frontier-substitution interpretation of Claude's zero interaction is a possible identification concern (zero interaction could also mean insensitivity or a null effect), but that is a statistical inference issue, not circularity.

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

The paper introduces no fitted constants; numerical treatments (audit probabilities, capability levels, sanction strengths) are experimental manipulations. The central claim rests primarily on the validity of the controlled comparisons and the representativeness of the constructed ecologies.

assumptions (3)
  • domain assumption The four-component institution model (rho, V, U, Phi) captures the relevant ways a collective turns private observations into public decision states.
    Section 2 defines the formal ecology and institution but does not prove that this decomposition is complete or that all institutional interventions fit within it.
  • ad hoc to paper Mechanism-breaking controls isolate the intended institutional signal without changing task difficulty or information quantity.
    Section 3 states every main comparison is paired, but the distributed-evidence below-threshold routing (3x2, 3x4) reduces the number of clue cards shown alongside routing completeness, so the coverage chain is partly entangled with total information.
  • domain assumption LLM agent behavior in these artificial ecologies is a valid model for studying collective reasoning failure.
    The paper explicitly disclaims human psychological equivalence (Section 1, Section 10) and relies on synthetic ecologies; the transfer of conclusions to real collective systems is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of When Do Institutions Beat Intelligence?." pith.science (2026). https://pith.science/paper/UNWSICY6

@misc{pith2026260811357,
  author       = {Pith},
  title        = {Pith review of: When Do Institutions Beat Intelligence?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UNWSICY6}},
  note         = {Machine review of arXiv:2608.11357}
}
read the original abstract

More capable agents do not necessarily form a more capable collective. A multi-agent system may jointly possess sufficient information yet fail because evidence is poorly routed, unreliable reports enter public belief, correlated claims masquerade as independent support, shared state becomes stale or strategically distorted, or useful evidence is exposed through an ineffective action interface. We ask when additional resources should improve the reasoner and when they should instead change the institutional structure through which the collective forms and acts on public information. Drawing on functional distinctions from research on group decision making and distributed cognition, we construct controlled artificial ecologies around four loci of collective failure: access and routing, admission and dependence, state maintenance and incentives, and representation and action. Across these ecologies, we separately vary model capability and institutional structure, pairing positive interventions with matched reasoning baselines and mechanism-breaking controls. The experiments reveal a consistent boundary: institutions help when they repair failures in how a collective constructs usable public state, but lose their advantage when their signals are uninformative or uncheckable, when stronger intelligence can perform the same transformation directly, or when the resulting state cannot support reliable action. Our results recast the choice between intelligence and institutions as a diagnosis of where collective reasoning fails.

Figures

Figures reproduced from arXiv: 2608.11357 by the authors.

Figure 1
Figure 1. Intelligence and institutions intervene at different levels of collective computation. Intelligence interventions [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Distributed evidence as an access bottleneck. Raw [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Structural information bottlenecks survive stronger reasoning over the same unusable state. The three ecologies share [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Checkability makes enforcement behaviorally [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Institutional effects track valid and behaviorally active signals. (a) Valid-rule gain minus broken-rule gain across [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Construction, substitution, and execution are dis [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: Capability, learned construction, and the action interface bound institutional advantage. (a) Frontier-model valid [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 23 canonical work pages

  1. [1]

    Amin Rahimian, Jason W

    Abdullah Almaatouq, M. Amin Rahimian, Jason W. Burton, and Abdullah J. Alhajri. 2022. The Distribution of Initial Estimates Moderates the Effect of Social Influence on the Wisdom of the Crowd.Scientific Reports12 (2022). doi:10.1038/ s41598-022-20551-7 8

    work page 2022

  2. [2]

    Joshua Becker, Devon Brackbill, and Damon Centola. 2017. Network Dynamics of Social Influence in the Wisdom of Crowds.Proceedings of the National Academy of Sciences114, 26 (2017), E5070–E5076. doi:10.1073/pnas.1615978114

    doi:10.1073/pnas.1615978114 2017

  3. [3]

    Guido Boella, Leendert Van Der Torre, and Harko Verhagen. 2006. Introduction to normative multiagent systems.Computational & Mathematical Organization Theory12, 2 (2006), 71–79

    work page 2006

  4. [4]

    Chi-Min Chan, Weize Chen, Yusheng Su, Jianxuan Yu, Wei Xue, Shanghang Zhang, and Jie Fu. 2024. ChatEval: Towards Better LLM-Based Evaluators through Multi-Agent Debate. InInternational Conference on Learning Representations

    work page 2024

  5. [5]

    Huan Chen, Xiang Song, Jian Jin, Pan Ren, and Liang-Jie Zhang. 2026. Toward an Organizational Science of Multi-Agent LLM Systems: Decoupling Who, How, and Which Algorithm.arXiv preprint arXiv:2607.25446(2026)

    work page Pith review arXiv 2026

  6. [6]

    Maxim Chupilkin. 2026. Artificial Institutions: How Institutional Design Shapes LLM Simulations.arXiv preprint arXiv:2608.04020(2026)

    work page Pith review arXiv 2026

  7. [7]

    Allan Dafoe et al. 2021. Cooperative AI: Machines Must Learn to Find Common Ground.Nature593 (2021), 33–36. doi:10.1038/d41586-021-01170-0

    doi:10.1038/d41586-021-01170-0 2021

  8. [8]

    Tenenbaum, and Igor Mor- datch

    Yilun Du, Shuang Li, Antonio Torralba, Joshua B. Tenenbaum, and Igor Mor- datch. 2024. Improving Factuality and Reasoning in Language Models through Multiagent Debate. InProceedings of the 41st International Conference on Machine Learning

    work page 2024

Show all 35 references
  1. [9]

    Marc Esteva, David De La Cruz, and Carles Sierra. 2002. ISLANDER: an electronic institutions editor. InProceedings of the first international joint conference on Autonomous agents and multiagent systems: part 3. 1045–1052

    2002

  2. [10]

    Marc Esteva, Juan-Antonio Rodriguez-Aguilar, Carles Sierra, Pere Garcia, and Josep L Arcos. 2001. On the formal specification of electronic institutions. InAgent Mediated Electronic Commerce: The European AgentLink Perspective. Springer, 126– 147

    2001

  3. [11]

    Chao Fei, Hongcheng Guo, and Yanghua Xiao. 2026. When Agents Evolve, Institutions Follow.arXiv preprint arXiv:2604.27691(2026)

    2026 arXiv

  4. [12]

    Vincenz Frey and Arnout van de Rijt. 2021. Social Influence Undermines the Wisdom of the Crowd in Sequential Decision Making.Management Science67, 7 (2021), 4273–4286. doi:10.1287/mnsc.2020.3713

    2021

  5. [13]

    Tobias Greitemeyer and Stefan Schulz-Hardt. 2003. Preference-Consistent Evalu- ation of Information in the Hidden Profile Paradigm.Journal of Personality and Social Psychology84, 2 (2003), 322–339. doi:10.1037/0022-3514.84.2.322

    2003 doi

  6. [14]

    Jean-Michel Hoc. 2000. Cognitive Aspects of Dynamic Situation Management. In Proceedings of the Human Factors and Ergonomics Society Annual Meeting, Vol. 44. 156–156. doi:10.1177/154193120004400141

    2000 doi

  7. [15]

    1995.Cognition in the Wild

    Edwin Hutchins. 1995.Cognition in the Wild. MIT Press

    1995

  8. [16]

    Xiao Liu et al. 2024. AgentBench: Evaluating LLMs as Agents. InInternational Conference on Learning Representations

    2024

  9. [17]

    Connie Yuan, and Poppy Lauretta McLeod

    Li Lu, Y. Connie Yuan, and Poppy Lauretta McLeod. 2012. Twenty-Five Years of Hidden Profiles in Group Decision Making: A Meta-Analysis.Personality and Social Psychology Review16, 1 (2012), 54–75. doi:10.1177/1088868311417243

    2012 doi

  10. [18]

    Kourken Michaelian and John Sutton. 2013. Distributed Cognition and Memory Research: History and Current Directions.Review of Philosophy and Psychology 4 (2013), 1–24. doi:10.1007/s13164-013-0131-x

    2013 doi

  11. [19]

    1990.Governing the Commons: The Evolution of Institutions for Collective Action

    Elinor Ostrom. 1990.Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. doi:10.1017/CBO9780511807763

    1990 doi

  12. [20]

    Elinor Ostrom. 2009. A General Framework for Analyzing Sustainability of Social- Ecological Systems.Science325, 5939 (2009), 419–422. doi:10.1126/science.1172133

    2009 doi

  13. [21]

    Bernstein

    Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S. Bernstein. 2023. Generative Agents: Interactive Simulacra of Human Behavior. InProceedings of the 36th Annual ACM Symposium on User Interface Software and Technology. doi:10.114...

    2023

  14. [22]

    Mike Scaife and Yvonne Rogers. 1996. External Cognition: How Do Graphical Representations Work?International Journal of Human-Computer Studies45, 2 (1996), 185–213. doi:10.1006/ijhc.1996.0048

    1996

  15. [23]

    Thomas Schillemans. 2022. Accountability and the Quality of Regulatory Judg- ment Processes.Public Performance & Management Review45 (2022), 473–498. doi:10.1080/15309576.2022.2040034

    2022

  16. [24]

    Garold Stasser and William Titus. 1985. Pooling of Unshared Information in Group Decision Making: Biased Information Sampling During Discussion.Journal of Personality and Social Psychology48, 6 (1985), 1467–1478. doi:10.1037/0022- 3514.48.6.1467

    1985 doi

  17. [25]

    Vaughan, and Dennis D

    Garold Stasser, Sandra I. Vaughan, and Dennis D. Stewart. 2000. Pooling Unshared Information: The Benefits of Knowing How Access to Information Is Distributed among Group Members.Organizational Behavior and Human Decision Processes 82, 1 (2000), 102–116. doi:10.1006/obhd.2000.2890

    2000

  18. [26]

    Marcantonio Bracale Syrnikov, Federico Pierucci, Marcello Galisai, Matteo Prandi, Piercosma Bisconti, Francesco Giarrusso, Olga Sorokoletova, Vincenzo Suriani, and Daniele Nardi. 2026. Institutional AI: Governing LLM collusion in multi-agent cournot markets via public governan...

    2026

  19. [27]

    Tetlock and Richard Boettger

    Philip E. Tetlock and Richard Boettger. 1989. Accountability: A Social Magnifier of the Dilution Effect.Journal of Personality and Social Psychology57, 3 (1989), 388–398. doi:10.1037/0022-3514.57.3.388

    1989 doi

  20. [28]

    Tetlock, Linda Skitka, and Richard Boettger

    Philip E. Tetlock, Linda Skitka, and Richard Boettger. 1989. Social and Cogni- tive Strategies for Coping with Accountability: Conformity, Complexity, and Bolstering.Journal of Personality and Social Psychology57, 4 (1989), 632–640. doi:10.1037/0022-3514.57.4.632

    1989 doi

  21. [29]

    Harsh Trivedi, Niranjan Balasubramanian, Tushar Khot, and Ashish Sabharwal

  22. [30]

    Rodríguez-Aguilar

    Paul Valckenaers, John Sauter, Carles Sierra, and Juan A. Rodríguez-Aguilar. 2007. Applications and Environments for Multi-Agent Systems.Autonomous Agents and Multi-Agent Systems14 (2007), 61–85. doi:10.1007/s10458-006-9002-5

    2007 doi

  23. [31]

    van Ginkel and Daan van Knippenberg

    Wendy P. van Ginkel and Daan van Knippenberg. 2009. Knowledge about the Distribution of Information and Group Decision Making: When and Why Does It Work?Organizational Behavior and Human Decision Processes108, 2 (2009), 218–229. doi:10.1016/j.obhdp.2008.10.003

    2009 doi

  24. [32]

    White, Doug Burger, and Chi Wang

    Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, Shaokun Zhang, Jiale Liu, Ahmed Hassan Awadallah, Ryen W. White, Doug Burger, and Chi Wang. 2024. AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation. InConfer...

    2024

  25. [33]

    Cohen, Ruslan Salakhutdinov, and Christopher D

    Zhilin Yang, Peng Qi, Saizheng Zhang, Yoshua Bengio, William W. Cohen, Ruslan Salakhutdinov, and Christopher D. Manning. 2018. HotpotQA: A Dataset for Diverse, Explainable Multi-Hop Question Answering. InProceedings of the 2018 Conference on Empirical Methods in Natural Langua...

    2018 doi

  26. [34]

    Jiajie Zhang and Donald A. Norman. 1994. Representations in Distributed Cogni- tive Tasks.Cognitive Science18, 1 (1994), 87–122. doi:10.1207/s15516709cog1801_3 9

    1994 doi

  27. [2022]

    Transactions of the Association for Computational Linguistics10 (2022), 539–554

    MuSiQue: Multihop Questions via Single-Hop Question Composition. Transactions of the Association for Computational Linguistics10 (2022), 539–554. doi:10.1162/tacl_a_00475

    2022 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.