Pith. sign in

REVIEW 3 major objections 4 minor 28 references

SkillShapley: Boundary-Adaptive Shapley Valuation for Skill Step Attribution in LLM Agents

T0 review · 3 major / 4 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 This paper claims that Shapley values are behaviorally meaningful for agent skill steps and that BAES approximates them efficiently from a small budget of skill evaluations.

desk verdict Solid Shapley-for-agent-skills idea with a cache-aware sampler, but the three-instance evaluation grid makes all empirical claims fragile. read the letter →

arxiv 2608.13173 v1 pith:MEEANQZO submitted 2026-08-13 cs.AI

Chang Liu , Yuqi Zhang , Yiman Zhong , Boyi Liu , Hengjun Wang , Shuyue Wei This is my paper · ORCID
classification cs.AI
keywords ShapleyvalueagentskillsLLMagentsstepattributionactivesamplingcoalitionalgametheorybudgetedapproximationBench
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

SkillShapley claims that the steps of an LLM agent skill can be valued as players in a cooperative game: each step's worth is its Shapley value, computed from benchmark success rates of skill variants that keep different subsets of steps. The paper argues these value rankings are behaviorally meaningful because deleting Shapley-top-ranked steps degrades task success faster than deleting steps ranked by isolated or leave-one-out scores, which produce many ties. Because exact Shapley requires exponential coalition evaluations, the paper introduces BAES, a budgeted active sampler that reuses cached one-flip comparisons to approximate the same rankings from far fewer unique skill variants. If correct, skill authors can get actionable step values from a few dozen runs instead of enumerating all subsets.

What carries the argument

BAES is the key mechanism: a cache-aware, budgeted active sampler for the size-stratified Shapley estimator $\hat\phi_i = \frac{1}{n}\sum_{k=0}^{n-1}\hat\mu_{i,k}$, where $\hat\mu_{i,k}$ is the empirical mean of observed one-flip marginal effects $\Delta_i(S)=v(S\cup\{i\})-v(S)$ in stratum $(i,k)$. It scores strata by $a_{i,k}=\sqrt{\hat\sigma^2_{i,k}+\epsilon}/\sqrt{m_{i,k}+1}$ so uncertain and under-sampled strata get priority, and acquires the unevaluated coalition maximizing $A(C)=\sum_{i:C\triangle\{i\}\in D} a_{i,k}\cdot b(v(C\triangle\{i\}))$ over cached one-flip neighbors. The warmup stage builds broad stratum coverage with anchors ($\emptyset$, all players, singletons, $(n-1)$-subsets) and a greedy cache-aware rule; the adaptive stage spends the remaining budget where cached edges reduce uncertainty most. The design exploits the paper's empirical observations that configuration evaluation dominates cost, rewards are discrete and high-variance, and rewards flatten in large coalition regions.

What would settle it

Pick one of the three skills, draw several different three-instance benchmark subsets (or enlarge to ten instances) and recompute exact Shapley rankings; if top-step identity or the relative error of BAES changes substantially, the reported attribution signal is an artifact of the instance triple.

Watch

Extended reading notes

Core claim

The central claim is that step-level Shapley values are a behaviorally meaningful attribution for agent skill steps and that the Boundary-Adaptive Edge Shapley (BAES) procedure approximates them efficiently. Treating each instruction block as a player and each retained-subset skill variant as a coalition, the paper defines the value of a step as its Shapley value with respect to the empirical success rate $v(S)$ on a fixed benchmark subset. Exact Shapley rankings, checked on three low-step-count SkillsBench tasks, produce differentiated step values and yield the fastest performance drop when top-ranked steps are removed, unlike Individual and Leave-One-Out scores that tie. BAES first warms up coverage across coalition sizes and then adaptively evaluates new coalitions chosen to form many reusable one-flip marginal edges with the cache, reporting that it reaches lower approximation error under smaller unique-configuration budgets than Monte Carlo, Quasi-Monte Carlo, paired Monte Carlo, and size-truncated Shapley baselines. The paper reads the resulting patterns as guidance for skill pruning, revision, and creation: high-value steps are procedural bridges connecting conditions to executable decisions, while low-value steps are locally correct but action-incomplete.

Load-bearing premise

Every reported value comes from evaluating coalitions on just three benchmark instances, so the exact-Shapley reference and BAES's error and removal curves all inherit whatever noise those three instances introduce.

Editorial extensions

If this is right

  • With a budget of about $3n^2$ unique skill variants, a practitioner can obtain step value rankings that track the exact Shapley ordering closely enough to guide pruning, inspection, and revision.
  • Deleting top-ranked Shapley steps causes a sharper drop in benchmark success than deleting steps by Individual, Leave-One-Out, or Random Removal, so the rankings carry behavioral meaning.
  • The same coalition cache can be reused for multiple attribution questions: one evaluation of a variant contributes one-flip edges to several strata, lowering the per-stratum cost.
  • Attribution patterns across tasks suggest skills should be written as compact, decision-complete guidance units, with procedural bridges made explicit and action-incomplete helper steps removed or templated.
  • Token cost of a skill variant is not proportional to coalition size; step-level editing is best seen as removing low-value content rather than guaranteed token savings.

Reading between the lines

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

  • If the three-instance evaluation is noisy, the stability of rankings across benchmark subsamples should be checked; a bootstrap over instances would tell whether BAES's advantage over Monte Carlo is robust to the choice of evaluation subset.
  • The boundary-adaptive acquisition rule may transfer to other valuation problems where evaluating a candidate is expensive and rewards are discrete, such as data subset valuation for few-shot prompts or test-suite minimization, though the paper does not claim this.
  • Because BAES is a biased approximation that optimizes ranking recovery, downstream users should re-run the removal curve after an edit to validate the edit rather than trusting a single value vector.
  • The paper's limitation about strongly coupled assembly-line workflows implies that a practical deployment should first verify that the coalition space is meaningfully executable before interpreting value differences.
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 / 4 minor

Summary. The paper formulates step-level attribution in LLM agent skills as a Shapley-value computation over skill steps, treating each instruction block as a player and benchmark success rate as the utility function. It proposes SkillShapley, which uses a two-phase budgeted estimator called BAES: a warmup stage that builds broad cache coverage over coalition sizes, followed by an adaptive stage that acquires coalitions maximizing reusable one-flip marginal edges. The empirical evaluation uses three SkillsBench skills (offer-letter-generator, manufacturing-fjsp-optimization, dialogue parser) with exact Shapley enumeration as the reference, reports top-removal validation curves, and compares BAES against Monte Carlo, Quasi-Monte Carlo, paired Monte Carlo, and size-k-truncated Shapley estimators under matched unique-configuration budgets. The central claims are that Shapley values are behaviorally meaningful for skill steps and that BAES approximates them more efficiently than existing estimators.

Significance. If the empirical claims are robust, the framework addresses a genuine gap: existing skill optimization methods treat skills as whole units, whereas SkillShapley provides per-step valuation with a practical budgeted estimator. The matched-budget comparison is a sound experimental design, and the cache-aware edge-reuse mechanism is a plausible way to reduce the number of expensive agent executions. The paper also gives credit to the problem by explicitly isolating the approximation question from the cost model. However, the evidence base is narrow: only three skills, each with a utility measured on three benchmark instances, and no stability analysis. The significance is therefore conditional on whether the coarse utility function is representative of the benchmark distribution.

major comments (3)
  1. [Appendix A; Eq. (3)] The utility v(S) is computed on exactly three benchmark instances, so v(S) takes values in {0, 1/3, 2/3, 1} and g = 1/3. This coarse, instance-specific realization is the 'exact Shapley' reference used in Section 5.2 and Figure 3. The paper provides no bootstrap, instance-subsample, or otherwise randomized stability analysis, despite acknowledging in Section 4.2 that rewards are 'discrete and noisy under limited repeated evaluation.' If the specific three-instance subset is not representative, the Shapley rankings, the removal curves, and the BAES approximation-error comparisons are all conditioned on that particular subset. This is load-bearing for both research questions, and the manuscript should add a stability analysis, for example by subsampling or bootstrapping over instances for at least one skill and reporting the sensitivity of the Shapley ranking and of BAES's relative error.
  2. [Figure 2] The removal validation in Figure 2 is not an independent confirmation of the Shapley rankings. The same v(S) from Eq. (3) is used both to compute the exact Shapley ranking and to measure the post-removal success rate (the mean of v over remaining coalitions). Under this protocol, a ranking that is overfit to the three-instance realization will automatically show a sharper drop than a more generalizable ranking. The validity claim would be substantially strengthened if the removal curves were evaluated on held-out benchmark instances that were not used to estimate the Shapley values, or at least on a bootstrap of the instances.
  3. [Section 5.3; RQ2] The stated research question RQ2 asks whether BAES can approximate exact Shapley rankings under a small budget, but the experiment in Figure 3 and the metrics in Appendix A report only player-wise MAE against the exact Shapley vector. No rank-oriented metric such as Kendall's tau, Spearman correlation, or top-k step overlap is reported. Since the paper's practical motivation is identifying high- and low-value steps rather than reporting numerically accurate Shapley scalars, the approximation claim would be better supported by reporting ranking-quality metrics alongside value error, particularly under the small budgets that are the intended use case.
minor comments (4)
  1. [Abstract] There is a typo in the abstract: 'i.e.,there' should be 'i.e., there'.
  2. [Figure 3 caption] The caption states the y-axis is 'error relative to full Shapley' but does not specify the metric; Appendix A says player-wise MAE is used. Please state that explicitly in the caption for self-containedness.
  3. [Section 5.1] The paper states that all model calls use temperature T = 0 but does not mention whether decoding is fully deterministic; if it is, the only source of stochasticity in v(S) is benchmark-instance selection, which reinforces the need for an instance-stability analysis in the major comments.
  4. [General] The paper does not state data or code availability; adding a reproducibility statement would improve the practical value of the method for practitioners who might want to apply SkillShapley to their own skills.

Circularity Check

1 steps flagged · score 3.0 of 10

BAES approximation is honestly benchmarked against exact Shapley, but the behavioral-validity claim (RQ1) uses a removal test on the same utility v(S) that defines Shapley, so it is a consistency check rather than an independent confirmation.

  1. self definitional [Section 5.2, Figure 2 (with Eq. (3) and Eq. (5))]
    "We evaluate attribution validity through top-ranked removal across all three exact-reference tasks (Figure 2). If a method truly identifies important steps, then deleting high-ranked steps should cause faster utility degradation than deleting low-ranked or randomly selected steps. In Figure 2, the full Shapley removal curves drop fastest across the exact-reference tasks, supporting the claim that Shapley values identify behaviorally important skill blocks more reliably than the simpler baselines."

    The removal curve's y-axis is 'mean success rate over all coalitions that do not contain the removed players,' i.e., averages of the same v(S) from Eq. (3) that enters the Shapley formula in Eq. (5). Shapley is a functional of v, so Figure 2 is a consistency check of Shapley against its own input rather than an independent behavioral measure. Since the paper itself asserts 'step interactions that are largely additive rather than synergistic' (Abstract), for additive games phi_i is the additive weight and top-removal degradation is mathematically forced; no external benchmark score, held-out instances, or human judgment validates the 'behaviorally meaningful' claim independently. The BAES efficiency claim is not circular because it is honestly measured against the exact Shapley reference.

full rationale

The central BAES claim is not circular: BAES estimates the standard Shapley value defined in Eq. (5), and its approximation error is measured against the exact Shapley vector computed from the same utility function, which is the correct external reference for an approximation-quality test. No fitted parameter is renamed as a prediction, and no load-bearing self-citation appears; SkillsBench [11] and other baselines are external. The only meaningful circularity is in the behavioral-validity experiment: the paper claims Shapley rankings are 'behaviorally meaningful' because removing top-ranked steps degrades utility, but that utility is exactly the v(S) from which Shapley was computed. This makes Figure 2 a self-consistency check, especially under the paper's own observation that step interactions are largely additive. The three-instance evaluation makes v coarse ({0,1/3,2/3,1}), which is a robustness/correctness limitation, not a circularity, and is acknowledged in the paper's limitation discussion. Overall, the derivation chain for BAES is self-contained; the partial circularity is confined to the validation of RQ1, giving a moderate score rather than a severe one.

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

The method itself uses the standard Shapley definition with no invented math, so the ledger is dominated by evaluative assumptions: the reward function on three instances is treated as the true utility, the manual segmentation defines the player set, and the additive-interaction observation justifies the sampling design. The hyperparameters of BAES are hand-set with no sensitivity analysis. No new entities are postulated.

free parameters (3)
  • Evaluation budget B = 3n^2 and warmup cap R = floor(0.4B) = (B,R) = (300,120), (243,97), (363,145) for the three skills
    The total unique-configuration budget and warmup cap are chosen without sensitivity analysis; the method's reported efficiency at 'low budget' is relative to this specific scaling, which is not derived.
  • Warmup and stopping hyperparameters = 60/85 strata thresholds, EMA half-life 3, top-30 strata, up to 3 coalitions per round, epsilon = (g/10)^2
    These constants control when warmup and adaptive phases stop and how acquisition is prioritized; no sensitivity or ablation is provided, so the estimator's behavior may depend on these hand-set values.
  • Manual segmentation of skill.md into blocks = n = 10, 9, 11 respectively
    The player set itself is chosen by human segmentation; different block boundaries would change all Shapley values. This is a modeling choice rather than a numeric fit, but it is a free choice affecting results.
assumptions (4)
  • standard math Shapley axioms (efficiency, symmetry, dummy, linearity) define the attribution target
    Eq. (5) is the textbook Shapley value; the paper adopts it without deriving or justifying why these axioms fit skill-step contributions.
  • domain assumption Skill performance defines a valid cooperative game over step subsets
    Section 3 assumes v(S) from Eq. (3) is a well-defined utility on all 2^n subsets with frontmatter and context fixed; this requires that step deletions never produce invalid or unrunnable skills, which the paper acknowledges is violated by strongly coupled assembly-line workflows (Section 6).
  • domain assumption Step interactions are largely additive; one-flip marginals capture the signal
    Section 4.2 states as an empirical observation that interactions are additive rather than synergistic and rewards flatten out; this motivates the boundary-adaptive design and stopping rule, but no interaction analysis or evidence is presented.
  • domain assumption Measured reward on three benchmark instances is a stable proxy for coalition value
    Appendix A: each configuration is evaluated on three instances, yielding values in {0, 1/3, 2/3, 1}; the paper treats exact Shapley computed from these noisy values as ground truth and provides no bootstrap or stability analysis.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SkillShapley: Boundary-Adaptive Shapley Valuation for Skill Step Attribution in LLM Agents." pith.science (2026). https://pith.science/paper/MEEANQZO

@misc{pith2026260813173,
  author       = {Pith},
  title        = {Pith review of: SkillShapley: Boundary-Adaptive Shapley Valuation for Skill Step Attribution in LLM Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MEEANQZO}},
  note         = {Machine review of arXiv:2608.13173}
}
read the original abstract

Agent skills are crucial external instructions that enable language agents to execute long procedural tasks such as coding or document processing. Existing agent skills are primarily created through human manual crafting or agent execution traces, with limited understanding of how each step contributes to overall skill performance on specific tasks; i.e., there remains an open problem in quantifying the contribution of individual steps within an agent skill. To address this issue, we first model skill-step attribution as a Shapley value-based contribution estimation problem, and then propose SkillShapley, a step-level attribution framework for agent skills. Notably, SkillShapley operates in two phases, motivated by key empirical insights, i.e., discretized benchmark rewards that create sharp performance cliffs, and step interactions that are largely additive rather than synergistic. Specifically, it first identifies informative coalitional regions, and then adaptively samples new coalitions that can yield reusable marginal evidence. Experiments on skills from the widely adopted SkillsBench demonstrate that our SkillShapley can effectively and efficiently identify high- or low-value skill steps, providing several key takeaways for agent skill creation.

Figures

Figures reproduced from arXiv: 2608.13173 by the authors.

Figure 1
Figure 1. Overview of BAES for budgeted skill-step attribution. The left panel shows how semantic [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Experiment 1 removal validation on the three SkillsBench tasks. Each panel corresponds [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Experiment 2 approximation error under increasing unique-configuration budgets. The x [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Token-cost diagnostic for Dialogue Parser coalitions. The y-axis shows the total token [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 20 canonical work pages

  1. [1]

    Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhari- wal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al

    Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhari- wal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. Language models are few-shot learners. InAdvances in Neural Information Processing Systems, volume 33, pages 1877–1901, 2020

    work page 1901

  2. [2]

    GPT-4 technical report, 2023

    OpenAI. GPT-4 technical report, 2023

    2023

  3. [3]

    Jimenez, Alexander Wettig, Kilian Lieret, Shunyu Yao, Karthik R

    John Yang, Carlos E. Jimenez, Alexander Wettig, Kilian Lieret, Shunyu Yao, Karthik R. Narasimhan, and Ofir Press. SWE-agent: Agent-computer interfaces enable automated soft- ware engineering. InThe Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024

    work page 2024

  4. [4]

    ReAct: Synergizing reasoning and acting in language models

    Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. ReAct: Synergizing reasoning and acting in language models. InInternational Conference on Learning Representations, 2023. doi: 10.48550/arXiv.2210.03629

    arXiv doi:10.48550/arxiv.2210.03629 2023

  5. [5]

    Joshi, Hanna Moazam, Heather 11 Miller, Matei Zaharia, and Christopher Potts

    Omar Khattab, Arnav Singhvi, Paridhi Maheshwari, Zhiyuan Zhang, Keshav Santhanam, Sri Vardhamanan, Saiful Haq, Ashutosh Sharma, Thomas T. Joshi, Hanna Moazam, Heather 11 Miller, Matei Zaharia, and Christopher Potts. DSPy: Compiling declarative language model calls into self-improving pipelines. InInternational Conference on Learning Representations,

    work page

  6. [6]

    Toolformer: Language models can teach themselves to use tools

    Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu, Maria Lomeli, Eric Ham- bro, Luke Zettlemoyer, Nicola Cancedda, and Thomas Scialom. Toolformer: Language models can teach themselves to use tools. InAdvances in Neural Information Processing Systems, volume 36, 2023. doi: 10.52202/075280-2997

    doi:10.52202/075280-2997 2023

  7. [7]

    Augmented language models: a survey.Transactions on Machine Learning Research, 2023

    Grégoire Mialon, Roberto Dessì, Maria Lomeli, Christoforos Nalmpantis, Ram Pasunuru, Roberta Raileanu, Baptiste Rozière, Timo Schick, Jane Dwivedi-Yu, Asli Celikyilmaz, et al. Augmented language models: a survey.Transactions on Machine Learning Research, 2023

    work page 2023

  8. [8]

    Lloyd S. Shapley. A value forn-person games. InContributions to the Theory of Games, volume 2, pages 307–317. Princeton University Press, 1953. doi: 10.1515/9781400881970-018

    doi:10.1515/9781400881970-018 1953

Show all 28 references
  1. [9]

    A unified approach to interpreting model predictions

    Scott M Lundberg and Su-In Lee. A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems, volume 30, 2017

    2017

  2. [10]

    Data Shapley: Equitable valuation of data for machine learning

    Amirata Ghorbani and James Zou. Data Shapley: Equitable valuation of data for machine learning. InProceedings of the 36th International Conference on Machine Learning, pages 2242–2251, 2019

    2019

  3. [11]

    SkillsBench: Benchmarking how well agent skills work across diverse tasks, 2026

    Xiangyi Li, Yimin Liu, Wenbo Chen, Bingran You, Zonglin Di, Yifeng He, Shenghan Zheng, Kyoung Whan Choe, Jiankai Sun, Shuyi Wang, et al. SkillsBench: Benchmarking how well agent skills work across diverse tasks, 2026

    2026

  4. [12]

    Skill set optimization: Reinforcing language model behavior via transferable skills

    KolbyNottingham,BodhisattwaMajumder,BhavanaDalviMishra,SameerSingh,PeterClark, and Roy Fox. Skill set optimization: Reinforcing language model behavior via transferable skills. InProceedings of the 41st International Conference on Machine Learning, 2024. doi: 10.48550/arXiv.2402.03241

    2024 doi arXiv

  5. [13]

    SkillAxe:SharpeningLLM-authored agent skills through evaluation-guided self-refinement, 2026

    SrishtiGautam,ArjunRadhakrishna,andSumitGulwani. SkillAxe:SharpeningLLM-authored agent skills through evaluation-guided self-refinement, 2026

    2026

  6. [14]

    SkillRe- ducer: Optimizing LLM agent skills for token efficiency, 2026

    Yudong Gao, Zongjie Li, Yuanyuan Yuan, Zimo Ji, Pingchuan Ma, and Shuai Wang. SkillRe- ducer: Optimizing LLM agent skills for token efficiency, 2026

    2026

  7. [15]

    SkCC: Portable and secure skill compilation for cross-framework LLM agents, 2026

    Yipeng Ouyang, Yi Xiao, Yuhao Gu, and Xianwei Zhang. SkCC: Portable and secure skill compilation for cross-framework LLM agents, 2026

    2026

  8. [16]

    Howcanweknowwhatlanguage models know?Transactions of the Association for Computational Linguistics, 8:423–438, 2020

    ZhengbaoJiang,FrankF.Xu,JunAraki,andGrahamNeubig. Howcanweknowwhatlanguage models know?Transactions of the Association for Computational Linguistics, 8:423–438, 2020. doi: 10.1162/tacl_a_00324

    2020 doi

  9. [17]

    LLMLingua: Com- pressing prompts for accelerated inference of large language models

    Huiqiang Jiang, Qianhui Wu, Chin-Yew Lin, Yuqing Yang, and Lili Qiu. LLMLingua: Com- pressing prompts for accelerated inference of large language models. InProceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 13358–13376, 2023. doi: 10....

    2023 doi

  10. [18]

    LongLLMLingua: Accelerating and enhancing LLMs in long context scenarios via prompt 12 compression

    Huiqiang Jiang, Qianhui Wu, Xufang Luo, Dongsheng Li, Chin-Yew Lin, Yuqing Yang, and Lili Qiu. LongLLMLingua: Accelerating and enhancing LLMs in long context scenarios via prompt 12 compression. InProceedings of the 62nd Annual Meeting of the Association for Computational Ling...

    2024 doi

  11. [19]

    SHAP- based explanation methods: A review for NLP interpretability

    Edoardo Mosca, Ferenc Szigeti, Stella Tragianni, Daniel Gallagher, and Georg Groh. SHAP- based explanation methods: A review for NLP interpretability. InProceedings of the 29th International Conference on Computational Linguistics, pages 4593–4603, 2022. doi: 10.18653/ v1/2022...

    2022

  12. [20]

    TokenSHAP: Interpreting large language models with Monte Carlo Shapley value estimation

    Miriam Horovicz and Roni Goldshmidt. TokenSHAP: Interpreting large language models with Monte Carlo Shapley value estimation. InProceedings of the 1st Workshop on NLP for Science (NLP4Science), pages 1–8, 2024. doi: 10.18653/v1/2024.emnlp-nlpscience.1

    2024 doi

  13. [21]

    Polynomial calculation of the Shapley value based on sampling.Computers & Operations Research, 36(5):1726–1730, 2009

    Javier Castro, Daniel Gómez, and Juan Tejada. Polynomial calculation of the Shapley value based on sampling.Computers & Operations Research, 36(5):1726–1730, 2009. doi: 10.1016/j. cor.2008.04.004

    2009 doi

  14. [22]

    Improving polynomial estima- tion of the Shapley value by stratified random sampling with optimum allocation.Computers & Operations Research, 82:180–188, 2017

    Javier Castro, Daniel Gómez, Elisenda Molina, and Juan Tejada. Improving polynomial estima- tion of the Shapley value by stratified random sampling with optimum allocation.Computers & Operations Research, 82:180–188, 2017. doi: 10.1016/j.cor.2016.12.013

    2017 doi

  15. [23]

    Burgess and Archie C

    Mark A. Burgess and Archie C. Chapman. Approximating the Shapley value using stratified empirical bernstein sampling. InProceedings of the Thirtieth International Joint Conference on Artificial Intelligence, pages 73–81, 2021. doi: 10.24963/ijcai.2021/11

    2021 doi

  16. [24]

    FastSHAP: Real-time Shapley value estimation

    Neil Jethani, Mukund Sudarshan, Ian Covert, Su-In Lee, and Rajesh Ranganath. FastSHAP: Real-time Shapley value estimation. InInternational Conference on Learning Representations, 2022

    2022

  17. [25]

    Covert, Scott M

    Hugh Chen, Ian C. Covert, Scott M. Lundberg, and Su-In Lee. Algorithms to estimate Shapley value feature attributions.Nature Machine Intelligence, 5(6):590–601, 2023. doi: 10.1038/ s42256-023-00657-x

    2023

  18. [26]

    Michael Maschler, Bezalel Peleg, and Lloyd S. Shapley. Geometric properties of the kernel, nucleolus, and related solution concepts.Mathematics of Operations Research, 4(4):303–338,

  19. [1979]

    doi: 10.1287/moor.4.4.303. 13 A Experiment Details For each evaluated skill, we manually segmentskill.mdinto semantically coherent instruction blocks,suchastask-entryguidance,decisionrules,APIexamples,validationinstructions,orcommon pitfalls. We do not split inside tightly cou...

    doi

  20. [2024]

    doi: 10.48550/arXiv.2310.03771

    doi arXiv

Pith tools

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