Pith. sign in

REVIEW 3 major objections 5 minor 57 references

xInv: Explainable Optimization of Inverse Problems

T0 review · 3 major / 5 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 An optimizer can explain itself if its simulator is instrumented to narrate forward and backward passes.

desk verdict A genuinely new pipeline for NL explanations of optimizer behavior with real utility evidence, but the faithfulness gap highlighted in Section 3.6 is real and should gate the central claim. read the letter →

arxiv 2506.11056 v2 pith:ABVCROIN submitted 2025-05-22 cs.LG cs.AI

Sean Memery , Kevin Denamganai , Anna Kapron-King , Kartic Subr This is my paper
classification cs.LGcs.AI
keywords inverseproblemsexplainableoptimizationdifferentiablesimulationnaturallanguageexplanationsmodelstraceshuman-in-the-loopinterpretabilityposthocexplanation
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

Inverse problems—estimating inputs so a known simulator produces a desired output—are usually solved by iterative optimization, but that optimization loop is hard for domain experts to read. This paper proposes xInv, a methodology that wraps a differentiable optimizer with a natural-language layer: the simulator emits events during its forward and backward passes, these events are converted into qualitative descriptions, and a language model assembles them into explanations. The authors argue that this simple instrumentation turns a cryptic optimization trace into a queryable narrative, and they demonstrate it on a railroad-trajectory design problem and on the training of a small language model. If correct, it gives domain experts a way to ask why an optimizer behaved as it did without reading raw numbers.

What carries the argument

The load-bearing mechanism is the three-phase trace pipeline: emission, transformation, and synthesis. At optimization step $k$ the optimizer emits events $E_k$, rewards $R_k$, and updates $U_k$; the transformation functions map those numerical quantities to qualitative natural-language descriptions whose magnitudes are scaled relative to the average change seen across the run; and a language model combines them first into per-step descriptions and then into a global narrative that includes reward-change information. Event and update sample rates control how fine-grained the trace is. The same instrumentation is applied to LM training by treating generated samples, losses, and per-layer parameter changes as the raw signals.

What would settle it

Take a set of optimization instances and rerun the question-answering and discrimination evaluations with the reward transformation $\Phi_R$ replaced by a constant mapping that never reports cost changes, while keeping the optimizer correct; if language models still identify the most beneficial obstacle removal at the accuracies reported in Section 3.3, the informative content is not actually coming from the trace, and if accuracy collapses, the assumption that the transformations must carry the information is confirmed.

Watch

Extended reading notes

Core claim

The paper's central claim is that the iterative optimization of inverse problems can be made explainable while keeping the underlying optimizer essentially traditional, by instrumenting it to emit signals. At each step the optimizer emits events, rewards, and parameter updates, and domain-defined transformation functions $\Phi_E$, $\Phi_R$, and $\Phi_U$ convert these numerical signals into natural-language statements such as "Change in acceleration: small (positive)" and "Control point 1: Magnitude: very small (ESE)." A language model then reads the resulting trace and produces step-level descriptions plus a global narrative, and users can ask follow-up questions interactively. The two demonstrations—a physics-based railroad optimization and the training of a small LM on physics question answering—are meant to show that the method works for both white-box and black-box systems.

Load-bearing premise

The paper assumes that the numerical-to-language transformation functions defined by a domain expert preserve the decision-relevant information; if they do not, the explanations can be uninformative or misleading even when the optimization itself is sound.

Editorial extensions

If this is right

  • A domain expert can converse with an optimizer: asking for a speed-prioritized versus cost-prioritized run, requesting state changes, and querying why one route is faster or cheaper.
  • Without changing the underlying optimizer, the generated explanations let a language model identify which obstacle removal would most improve the result, with accuracy well above random chance.
  • Explanations carry enough information for a language model to distinguish optimized control points from noisy distractors even at low noise levels, with performance approaching the upper bound once noise grows.
  • Human participants rate the explanation as matching the true optimization significantly more than a distractor, and the preference strengthens for respondents who take longer to read.
  • Because only the emission and transformation functions are domain-specific, the same method transfers from a physics simulator to a black-box neural-network training run.

Reading between the lines

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

  • If this pipeline is sound, the practical bottleneck shifts from optimizer design to trace design: the fidelity of any explanation is bounded by what the domain expert chooses to expose through $\Phi_E$, $\Phi_R$, and $\Phi_U$.
  • The same idea could be used as an audit log for high-stakes optimization decisions, since it converts a numerical history into a reviewable narrative that a human can inspect after the fact.
  • A natural stress test would be to run the obstacle-identification task with deliberately impoverished transformation functions to see how much information loss the language model can tolerate before decisions degrade.
  • The method is likely to compose with other explainability tools, making the natural-language trace a reusable interface for attribution maps, sensitivity analysis, or policy-capturing techniques.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes xInv, a methodology for instrumenting differentiable simulators and optimizers so that they emit natural-language events, rewards, and parameter-update descriptions during forward and backward passes, and then uses a language model in a post-processing step to synthesize step-level and global explanations. The approach is demonstrated on a railroad-trajectory optimization problem with several standard optimizers, and on a small language-model training run, with evaluations consisting of an obstacle-removal question-answering task, a control-point discrimination task, and a human user study.

Significance. If the central claim holds, the paper offers a simple, modular bridge between numerical optimization and human-interpretable, queryable explanations, with broad potential applicability in scientific and engineering domains. The paper has clear strengths: it ships a concrete instantiation with a conversational agent, it evaluates across multiple LMs and with human raters, and the control-point discrimination and user-study results indicate that the generated descriptions carry usable information. However, the evidence currently supports usefulness and plausibility of explanations, not faithfulness to the underlying trace, and the LM-training example contains an unsupported causal attribution. The central idea is worth publishing once the faithfulness gap is addressed or the claims are appropriately scoped.

major comments (3)
  1. [Section 3.6] The stepwise example states: "This qualitative leap was directly attributable to improved contextual integration and normalization, as highlighted by our framework's attribution maps." The method described in Section 2.2 and the supplemental logs record only validation samples, loss differences, and average per-layer parameter-magnitude changes; no attribution maps, gradients, or ablations are defined anywhere in the manuscript. This causal assertion is therefore not grounded in the logged trace, and it is exactly the kind of unsupported statement that could mislead a domain expert. Please either remove the causal and attribution language, or add a genuine attribution mechanism (e.g., per-layer gradient norms or counterfactual ablations) and verify that the explanation is supported by it.
  2. [Sections 3.3-3.5] The three main evaluations are utility and discrimination tests: above-chance obstacle identification, target-versus-distractor matching, and human rating preference for the target visualization. Each of these can be passed by explanations that are plausible but not faithful to the logged trace, because a wrong explanation that is correlated with the correct one will still separate targets from distractors. Since the paper's central claim is about explainability, at least one evaluation should directly check faithfulness, for example by asking an LM or a human judge to verify individual explanation statements against the logged events, rewards, and updates, or by injecting an event that contradicts the explanation and measuring whether it is detected.
  3. [Section 3.3] The question-answering evaluation lacks a baseline in which the LM is given the raw numerical trace (events, rewards, updates) instead of the natural-language transformation. As a result, the observed above-chance accuracy could be driven by the LM's prior knowledge or by the initial path events, rather than by the proposed natural-language explanation. The Numerical condition in Section 3.4 covers only control-point positions, not the full numerical trace. Please add a raw-trace condition, or an ablation that removes the natural-language descriptions, to isolate the contribution of the proposed transformation.
minor comments (5)
  1. [Section 2.1, Eq. (2)] The formula t_m = (-v_m + sqrt(v_m^2 + 2 a_m s_m))/a_m is undefined when a_m = 0; please state the limiting case or otherwise handle this edge case in the simulation description.
  2. [Section 3.5] The user study reports a "statistically significant preference" for the target variant, but no test statistic, p-value, or effect size is reported; please specify the statistical test and the corresponding values.
  3. [Section 4, first bullet] The discussion states that "Section 3.1 demonstrates how LMs successfully interact with optimization systems," but Section 3.1 is the numerical optimization validation; the conversational interaction is in Section 3.2. Please correct the cross-reference.
  4. [Section 2.2 and Algorithm 1] The notation is inconsistent: Algorithm 1 uses "amu" and "aeta" for the frictional and air-resistance accelerations, while the text uses a_mu and a_eta; also the definition of E_k in Eq. (3) mixes superscripts and subscripts. Please unify the notation.
  5. [Supplemental Material] The supplemental material repeatedly refers to "Section 4.2" for details, but Section 4.2 in the main paper is the Conclusion; these references should point to the actual methodological sections.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: xInv's explanations are validated against external targets, and its only self-citation is a non-load-bearing related-work reference.

full rationale

The paper's chain is a pipeline, not a fitted prediction: it instruments an optimizer to emit events/rewards/updates, applies user-defined transformations Phi_E/Phi_R/Phi_U, and has an LM synthesize explanations. No parameter is fitted to the explanation target and no evaluated quantity is defined in terms of the explanation. The quantitative evaluations use external ground truths: Section 3.3 computes obstacle-removal effects by re-running simulations; Section 3.4 tests discrimination against noisy distractor control points, requiring LM inference from qualitative descriptions (the Numerical condition is explicitly an upper bound, not the explanation being tested); Section 3.5 uses human ratings of text-to-visualization match. The only overlapping-author citation, Memery et al. [37], appears in related work as a limitation reference and is not load-bearing. The main caveat is Section 3.6, where the claim that a qualitative leap was 'directly attributable to improved contextual integration and normalization, as highlighted by our framework's attribution maps' is not supported by the logged trace (average per-layer parameter changes and text samples) and no attribution maps are defined; this is an unsupported faithfulness/correctness claim, not a circular derivation, because the explanation goes beyond its inputs rather than being equivalent to them. Section 4.1 explicitly acknowledges the method's dependence on user-defined transformations. No equation or prediction reduces to its own inputs by construction.

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

The central claim rests on domain assumptions about information preservation in qualitative traces and LM reliability, plus the practical assumption that transformation functions can be authored per domain. No fitted free parameters or invented entities are introduced; the quantitative experiments are empirical demonstrations rather than derivations.

assumptions (3)
  • domain assumption Qualitative natural-language traces (events, rewards, updates) preserve sufficient decision-relevant information about the optimization process.
    The method deliberately replaces exact numerical values with qualitative descriptors in Section 2.2; if this quantization discards the information needed to explain decisions, the explanations lose value. The paper's evaluations support this empirically but do not prove it.
  • domain assumption Language models can generate faithful and useful explanations from these qualitative traces.
    The post-hoc explanation phase in Section 2.3 relies on LM competence in summarization and reasoning; the tests show LMs can use the descriptions, but faithfulness of the generated narratives to the optimizer's true behavior is not directly measured.
  • domain assumption Domain experts can specify the transformation functions Phi_E, Phi_R, Phi_U for a new domain.
    The authors state this explicitly as a limitation in Section 4.1: 'Our method depends on user-defined events, rewards, and updates, which might be challenging in some domains.'

how reviews work

0 comments
Cite this review

Pith. "Pith review of xInv: Explainable Optimization of Inverse Problems." pith.science (2026). https://pith.science/paper/ABVCROIN

@misc{pith2026250611056,
  author       = {Pith},
  title        = {Pith review of: xInv: Explainable Optimization of Inverse Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ABVCROIN}},
  note         = {Machine review of arXiv:2506.11056}
}
read the original abstract

Inverse problems are central to a wide range of fields, including healthcare, climate science, and agriculture. They involve the estimation of inputs, typically via iterative optimization, to some known forward model so that it produces a desired outcome. Despite considerable development in the explainability and interpretability of forward models, the iterative optimization of inverse problems remains largely cryptic to domain experts. We propose a methodology to produce explanations, from traces produced by an optimizer, that are interpretable by humans at the abstraction of the domain. The central idea in our approach is to instrument a differentiable simulator so that it emits natural language events during its forward and backward passes. In a post-process, we use a Language Model to create an explanation from the list of events. We demonstrate the effectiveness of our approach with an illustrative optimization problem and an example involving the training of a neural network.

Figures

Figures reproduced from arXiv: 2506.11056 by the authors.

Figure 1
Figure 1. We instrument traditional inverse-solving pipelines (a) to emit two natural language signals, [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Comparing relative savings in time (a) and cost (b) for different optimizers. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Question-answering well over chance. Different LMs were provided one of three descrip [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Success rates of Llama-4-Maverick[2] at the discrimination tasks, for different values of the standard deviation σ parameterising the distractor generation. 1 2 3 4 5 Rating 0 20 40 60 80 100 Count Distractor Target Low Medium High Time Taken 1 2 3 4 5 Median Rating Di…
Figure 5
Figure 5. Figure 5: Our user study. Left: a histogram of the ratings for our explanations against the correct [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Success rates of each language model at the discrimination tasks, for different values of [PITH_FULL_IMAGE:figures/full_fig_p030_6.png]
Figure 7
Figure 7. Figure 7: Explanation of task shown to participants. [PITH_FULL_IMAGE:figures/full_fig_p038_7.png]
Figure 8
Figure 8. Figure 8: Example of a question shown to participants. [PITH_FULL_IMAGE:figures/full_fig_p039_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

57 extracted references · 38 canonical work pages

  1. [1]

    Llama 3 model card

    AI@Meta. Llama 3 model card. 2024

    2024

  2. [2]

    Llama 4 model card

    AI@Meta. Llama 4 model card. 2025

    work page 2025

  3. [3]

    Karakasidis

    Dimitrios Angelis, Filippos Sofos, and Theodoros E. Karakasidis. Artificial intelligence in physical sciences: Symbolic regression trends and perspectives. Archives of Computational Methods in Engineering, pages 1 – 21, 2023

    work page 2023

  4. [4]

    Physical reasoning in infancy

    Renee Baillargeon. Physical reasoning in infancy. The cognitive neurosciences, pages 181–204, 1995

    work page 1995

  5. [5]

    Object permanence in five- month-old infants

    Renee Baillargeon, Elizabeth S Spelke, and Stanley Wasserman. Object permanence in five- month-old infants. Cognition, 20(3):191–208, 1985

    work page 1985

  6. [6]

    Krishnan

    Vahid Balazadeh, Mohammadmehdi Ataei, Hyunmin Cheong, Amir Hosein Khasahmadi, and Rahul G. Krishnan. Synthetic vision: Training vision-language models to understand physics, 2024

    work page 2024

  7. [7]

    George E. P. Box and Kenneth B. Wilson. On the experimental attainment of optimum conditions. Journal of the royal statistical society series b-methodological, 13:1–38, 1951

    work page 1951

  8. [8]

    Explainable bayesian optimization, 2025

    Tanmay Chakraborty, Christian Wirth, and Christin Seifert. Explainable bayesian optimization, 2025

    work page 2025

Show all 57 references
  1. [9]

    Llmphy: Complex physical reasoning using large language models and world models, 2024

    Anoop Cherian, Radu Corcodel, Siddarth Jain, and Diego Romeres. Llmphy: Complex physical reasoning using large language models and world models, 2024

    2024

  2. [10]

    Large language models as faithful explainers.ArXiv, abs/2402.04678, 2024

    Yu-Neng Chuang, Guanchu Wang, Chia yuan Chang, Ruixiang Tang, Fan Yang, Mengnan Du, Xuanting Cai, and Xia Hu. Large language models as faithful explainers.ArXiv, abs/2402.04678, 2024

    2024

  3. [11]

    Training verifiers to solve math word problems

    Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168, 2021. 34

    arXiv

  4. [12]

    Craven and Jude W

    Mark W. Craven and Jude W. Shavlik. Extracting tree-structured representations of trained networks. In Proceedings of the 9th International Conference on Neural Information Processing Systems, NIPS’95, page 24–30, Cambridge, MA, USA, 1995. MIT Press

    1995

  5. [13]

    Ai-assisted schedule explainer for nurse rostering

    Kristijonas Cyras, Amin Karamlou, Myles Lee, Dimitrios Letsios, Ruth Misener, and Francesca Toni. Ai-assisted schedule explainer for nurse rostering. InProceedings of the 19th International Conference on Autonomous Agents and MultiAgent Systems, AAMAS ’20, page 2101–2103, Rich...

    2020

  6. [14]

    Using large language model to solve and explain physics word problems approaching human level

    Jingzhe Ding, Yan Cen, and Xinyuan Wei. Using large language model to solve and explain physics word problems approaching human level. arXiv preprint arXiv:2309.08182, 2023

    2023 arXiv

  7. [15]

    Upol Ehsan and Mark O. Riedl. Explainability pitfalls: Beyond dark patterns in explainable ai, 2021

    2021

  8. [16]

    Minedojo: Building open-ended embodied agents with internet-scale knowledge, 2022

    Linxi Fan, Guanzhi Wang, Yunfan Jiang, Ajay Mandlekar, Yuncong Yang, Haoyi Zhu, Andrew Tang, De-An Huang, Yuke Zhu, and Anima Anandkumar. Minedojo: Building open-ended embodied agents with internet-scale knowledge, 2022

    2022

  9. [17]

    Gállego, and Marta Ruiz Costa-jussà

    Javier Ferrando, Gerard I. Gállego, and Marta Ruiz Costa-jussà. Measuring the mixing of contextual information in the transformer. ArXiv, abs/2203.04212, 2022

    2022 arXiv

  10. [18]

    A rule-based view of query optimization

    Johann Christoph Freytag. A rule-based view of query optimization. SIGMOD Rec. , 16(3):173–180, December 1987

    1987

  11. [19]

    Bias and fairness in large language models: A survey

    Isabel O Gallegos, Ryan A Rossi, Joe Barrow, Md Mehrab Tanjim, Sungchul Kim, Franck Dernoncourt, Tong Yu, Ruiyi Zhang, and Nesreen K Ahmed. Bias and fairness in large language models: A survey. Computational Linguistics, pages 1–79, 2024

    2024

  12. [20]

    Explainable ai via learning to optimize

    Howard Heaton and Samy Wu Fung. Explainable ai via learning to optimize. Scientific Reports, 13(1):10103, 2023

    2023

  13. [21]

    Measuring mathematical problem solving with the math dataset

    Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. Measuring mathematical problem solving with the math dataset. arXiv preprint arXiv:2103.03874, 2021

    2021 arXiv

  14. [22]

    Rothkopf

    Inga Ibs, Claire Ott, Frank Jäkel, and Constantin A. Rothkopf. From human explanations to explainable ai: Insights from constrained optimization. Cognitive Systems Research, 88:101297, 2024

    2024

  15. [23]

    Grasp: A novel benchmark for evaluating language grounding and situated physics understanding in multimodal language models

    Serwan Jassim, Mario Holubar, Annika Richter, Cornelius Wolff, Xenia Ohmer, and Elia Bruni. Grasp: A novel benchmark for evaluating language grounding and situated physics understanding in multimodal language models. In Kate Larson, editor, Proceedings of the Thirty-Third Inte...

  16. [24]

    Explainable reinforcement learning via reward decomposition

    Zoe Juozapaitis, Anurag Koul, Alan Fern, Martin Erwig, and Finale Doshi-Velez. Explainable reinforcement learning via reward decomposition. 2019

    2019

  17. [25]

    Llms can’t plan, but can help planning in llm-modulo frameworks

    Subbarao Kambhampati, Karthik Valmeekam, Lin Guan, Kaya Stechly, Mudit Verma, Siddhant Bhambri, Lucas Saldyt, and Anil Murthy. Llms can’t plan, but can help planning in llm-modulo frameworks. arXiv preprint arXiv:2402.01817, 2024

    2024 arXiv

  18. [26]

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

    Omar Khattab, Arnav Singhvi, Paridhi Maheshwari, Zhiyuan Zhang, Keshav Santhanam, Sri Vardhamanan A, Saiful Haq, Ashutosh Sharma, Thomas T. Joshi, Hanna Moazam, Heather Miller, Matei Zaharia, and Christopher Potts. DSPy: Compiling declarative language model calls into state-of...

    2024

  19. [27]

    Perception and understanding of effects of gravity and inertia on object motion

    In-Kyeong Kim and Elizabeth S Spelke. Perception and understanding of effects of gravity and inertia on object motion. Developmental Science, 2(3):339–362, 1999. 35

    1999

  20. [28]

    Measuring faithfulness in chain-of-thought reasoning

    Tamera Lanham, Anna Chen, Ansh Radhakrishnan, Benoit Steiner, Carson Denison, Danny Hernandez, Dustin Li, Esin Durmus, Evan Hubinger, Jackson Kernion, et al. Measuring faithfulness in chain-of-thought reasoning. arXiv preprint arXiv:2307.13702, 2023

    2023 arXiv

  21. [29]

    Ruibo Liu, Jason Wei, Shixiang Shane Gu, Te-Yen Wu, Soroush V osoughi, Claire Cui, Denny Zhou, and Andrew M. Dai. Mind’s eye: Grounded language model reasoning through simulation. ArXiv, abs/2210.05359, 2022

    2022 arXiv

  22. [30]

    Improving the interpretability of deep neural networks with knowledge distillation, 2018

    Xuan Liu, Xiaoguang Wang, and Stan Matwin. Improving the interpretability of deep neural networks with knowledge distillation, 2018

    2018

  23. [31]

    Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts

    Pan Lu, Hritik Bansal, Tony Xia, Jiacheng Liu, Chunyuan Li, Hannaneh Hajishirzi, Hao Cheng, Kai-Wei Chang, Michel Galley, and Jianfeng Gao. Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts. In The Twelfth International Conference on Learning...

    2024

  24. [32]

    A unified approach to interpreting model predictions, 2017

    Scott Lundberg and Su-In Lee. A unified approach to interpreting model predictions, 2017

    2017

  25. [33]

    Lundberg and Su-In Lee

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

    2017

  26. [34]

    From understanding to utilization: A survey on explainability for large language models

    Haoyan Luo and Lucia Specia. From understanding to utilization: A survey on explainability for large language models. ArXiv, abs/2401.12874, 2024

    2024 arXiv

  27. [35]

    Adapting while learning: Grounding llms for scientific problems with intelligent tool usage adaptation

    Bohan Lyu, Yadi Cao, Duncan Watson-Parris, Leon Bergen, Taylor Berg-Kirkpatrick, and Rose Yu. Adapting while learning: Grounding llms for scientific problems with intelligent tool usage adaptation. ArXiv, abs/2411.00412, 2024

    2024 arXiv

  28. [36]

    Mecattaf, Ben Slater, Marko Tevsi’c, Jonathan Prunty, Konstantinos V oudouris, and Lucy G

    Matteo G. Mecattaf, Ben Slater, Marko Tevsi’c, Jonathan Prunty, Konstantinos V oudouris, and Lucy G. Cheke. A little less conversation, a little more action, please: Investigating the physical common-sense of llms in a 3d embodied environment. 2024

    2024

  29. [37]

    Simlm: Can language models infer parameters of physical systems?, 2024

    Sean Memery, Mirella Lapata, and Kartic Subr. Simlm: Can language models infer parameters of physical systems?, 2024

    2024

  30. [38]

    Explainable reinforcement learning: A survey and comparative review

    Stephanie Milani, Nicholay Topin, Manuela Veloso, and Fei Fang. Explainable reinforcement learning: A survey and comparative review. ACM Computing Surveys, 56:1 – 36, 2023

    2023

  31. [39]

    Myers and Douglas C

    Raymond H. Myers and Douglas C. Montgomery. Response Surface Methodology: Process and Product in Optimization Using Designed Experiments. John Wiley & Sons, Inc., USA, 1st edition, 1995

    1995

  32. [40]

    Gpt-4 technical report

    OpenAI. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023

    2023 arXiv

  33. [41]

    Erika Puiutta and Eric M. S. P. Veith. Explainable reinforcement learning: A survey. ArXiv, abs/2005.06247, 2020

    2005 arXiv

  34. [42]

    Explainable distributed constraint optimization problems, 2025

    Ben Rachmut, Stylianos Loukas Vasileiou, Nimrod Meir Weinstein, Roie Zivan, and William Yeoh. Explainable distributed constraint optimization problems, 2025

    2025

  35. [43]

    Explainable reinforce- ment learning via temporal policy decomposition, 2025

    Franco Ruggeri, Alessio Russo, Rafia Inam, and Karl Henrik Johansson. Explainable reinforce- ment learning via temporal policy decomposition, 2025

    2025

  36. [44]

    Andrea Saltelli, Stefano Tarantola, Francesca Campolongo, and M. Ratto. Sensitivity analysis in practice: A guide to assessing scientific models. 06 2004

    2004

  37. [45]

    Origins of knowledge

    Elizabeth S Spelke, Karen Breinlinger, Janet Macomber, and Kristen Jacobson. Origins of knowledge. Psychological review, 99(4):605, 1992

    1992

  38. [46]

    Axiomatic attribution for deep networks

    Mukund Sundararajan, Ankur Taly, and Qiqi Yan. Axiomatic attribution for deep networks. In International Conference on Machine Learning, 2017

    2017

  39. [47]

    Mujoco: A physics engine for model-based control

    Emanuel Todorov, Tom Erez, and Yuval Tassa. Mujoco: A physics engine for model-based control. In 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 5026–5033. IEEE, 2012. 36

    2012

  40. [48]

    Language models don’t always say what they think: unfaithful explanations in chain-of-thought prompting

    Miles Turpin, Julian Michael, Ethan Perez, and Samuel Bowman. Language models don’t always say what they think: unfaithful explanations in chain-of-thought prompting. Advances in Neural Information Processing Systems, 36, 2024

    2024

  41. [49]

    Evaluating object permanence in embodied agents using the animal-ai environment

    Konstantinos V oudouris, Niall Donnelly, Danaja Rutar, Ryan Burnell, John Burden, José Hernández-Orallo, and Lucy G Cheke. Evaluating object permanence in embodied agents using the animal-ai environment. https://ceur-ws. org/Vol-3169/paper2. pdf, 2022

    2022

  42. [50]

    V oyager: An open-ended embodied agent with large language models

    Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu, Linxi (Jim) Fan, and Anima Anandkumar. V oyager: An open-ended embodied agent with large language models. Trans. Mach. Learn. Res., 2024, 2023

    2024

  43. [51]

    Chain-of-thought prompting elicits reasoning in large language models

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824–24837, 2022

    2022

  44. [52]

    ggplot2: Elegant Graphics for Data Analysis

    Hadley Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer Publishing Company, Incorporated, 2nd edition, 2016

    2016

  45. [53]

    Number cookbook: Num- ber understanding of language models and how to improve it

    Haotong Yang, Yi Hu, Shijia Kang, Zhouchen Lin, and Muhan Zhang. Number cookbook: Num- ber understanding of language models and how to improve it. In The Thirteenth International Conference on Learning Representations, 2025

    2025

  46. [54]

    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. In International Conference on Learning Representations (ICLR), 2023

    2023

  47. [55]

    Yates, Edward C

    William B. Yates, Edward C. Keedwell, and Ahmed Kheiri. Explainable optimisation through online and offline hyper-heuristics. ACM Trans. Evol. Learn. Optim. , October 2024. Just Accepted

    2024

  48. [56]

    Explainability for large language models: A survey, 2023

    Haiyan Zhao, Hanjie Chen, Fan Yang, Ninghao Liu, Huiqi Deng, Hengyi Cai, Shuaiqiang Wang, Dawei Yin, and Mengnan Du. Explainability for large language models: A survey, 2023. 37 Figure 7: Explanation of task shown to participants. 38 Figure 8: Example of a question shown to pa...

    2023

  49. [6305]

    Main Track

    International Joint Conferences on Artificial Intelligence Organization, 8 2024. Main Track

    2024

Pith tools

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