mlsanityData to 7 Oct 2026

Agentic-imodels: Evolving agentic interpretability tools via autoresearch

arXiv 2605.03808 ↗First posted 5 May 2026cs.AI cs.CL cs.LGcited by 1

On the map: LLM agents and tool use → Web, search, and research agents

Abstract (arXiv)

Agentic data science (ADS) systems are rapidly improving their capability to autonomously analyze, fit, and interpret data, potentially moving towards a future where agents conduct the vast majority of data-science work. However, current ADS systems use statistical tools designed to be interpretable by humans, rather than interpretable by agents. To address this, we introduce Agentic-imodels, an agentic autoresearch loop that evolves data-science tools designed to be interpretable by agents. Specifically, it develops a library of scikit-learn-compatible regressors for tabular data that are optimized for both predictive performance and a novel LLM-based interpretability metric. The metric measures a suite of LLM-graded tests that probe whether a fitted model's string representation is "simulatable" by an LLM, i.e. whether the LLM can answer questions about the model's behavior by reading its string output alone. We find that the evolved models jointly improve predictive performance and agent-facing interpretability, generalizing to new datasets and new interpretability tests. Furthermore, these evolved models improve downstream end-to-end ADS, increasing performance for Copilot CLI, Claude Code, and Codex on the BLADE benchmark by up to 73%

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use. Builds on 55 of its 81 references that are in the map. First posted on arXiv 5 May 2026.

Builds on

55 of 81 references.

Closest ancestors · 5

Same direction, newest first

  1. 2026DSGym: A Holistic Framework for Evaluating and Training Data Science Agents22 Jan
  2. 2025Agent Laboratory: Using LLM Agents as Research Assistants8 Jan
  3. 2024ScienceAgentBench: Toward Rigorous Assessment of Language Agents for Data-Driven Scientific Discovery7 Oct
  4. 2024BLADE: Benchmarking Language Model Agents for Data-Driven Science19 Aug
  5. 2024DS-Agent: Automated Data Science by Empowering Large Language Models with Case-Based Reasoning27 Feb

Other references · 50

  1. 2024GPT-4o System Cardcited by 4,467
  2. 2024The AI Scientist: Towards Fully Automated Open-Ended Scientific Discoverycited by 1,002
  3. 2025AlphaEvolve: A coding agent for scientific and algorithmic discoverycited by 763
  4. 2016The Mythos of Model Interpretabilitycited by 716
  5. 2020Does the Whole Exceed its Parts? The Effect of AI Explanations on Complementary Team Performancecited by 396
  6. 2021Interpretable Machine Learning: Fundamental Principles and 10 Grand Challengescited by 364
  7. 2019Interpretable machine learning: definitions, methods, and applicationscited by 303
  8. 2025Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agentscited by 294
  9. 2026Meta-Harness: End-to-End Optimization of Model Harnessescited by 248
  10. 2019InterpretML: A Unified Framework for Machine Learning Interpretabilitycited by 243
  11. 2022Measuring Progress on Scalable Oversight for Large Language Modelscited by 212
  12. 2025TabArena: A Living Benchmark for Machine Learning on Tabular Datacited by 152
38 more
  1. 2025Dynamic Cheatsheet: Test-Time Learning with Adaptive Memorycited by 146
  2. 2015Interpretable classifiers using rules and Bayesian analysis: Building a better stroke prediction modelcited by 140
  3. 2017PMLB: A Large Benchmark Suite for Machine Learning Evaluation and Comparisoncited by 139
  4. 2025Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memorycited by 129
  5. 2023Large Language Models for Automated Open-domain Scientific Hypotheses Discoverycited by 100
  6. 2026EvoSkill: Automated Skill Discovery for Multi-Agent Systemscited by 96
  7. 2026Learning to Discover at Test Timecited by 94
  8. 2015Supersparse Linear Integer Models for Optimized Medical Scoring Systemscited by 76
  9. 2026SkillClaw: Let Skills Evolve Collectively with Agentic Evolvercited by 76
  10. 2020Human Factors in Model Interpretability: Industry Practices, Challenges, and Needscited by 73
  11. 2019An Evaluation of the Human-Interpretability of Explanationcited by 71
  12. 2024A Multimodal Automated Interpretability Agentcited by 62
  13. 2023Goal Driven Discovery of Distributional Differences via Language Descriptionscited by 61
  14. 2023Explaining black box text modules in natural language with language modelscited by 61
  15. 2026Memento-Skills: Let Agents Design Agentscited by 56
  16. 2025ThetaEvolve: Test-time Learning on Open Problemscited by 52
  17. 2025Flipping the Dialogue: Training and Evaluating User Language Modelscited by 50
  18. 2025Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotationcited by 43
  19. 2018Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Insteadcited by 40
  20. 2025Evaluating Large Language Models in Scientific Discoverycited by 38
  21. 2026SKILLFOUNDRY: Building Self-Evolving Agent Skill Libraries from Heterogeneous Scientific Resourcescited by 38
  22. 2022Augmenting Interpretable Models with LLMs during Trainingcited by 36
  23. 2026HumanLM: Simulating Users with State Alignment Beats Response Imitationcited by 31
  24. 2025The More You Automate, the Less You See: Hidden Pitfalls of AI Scientist Systemscited by 24
  25. 2022Hierarchical Shrinkage: improving the accuracy and interpretability of tree-based methodscited by 18
  26. 2022GSCLIP : A Framework for Explaining Distribution Shifts in Natural Languagecited by 14
  27. 2025AI Safety for Everyonecited by 14
  28. 2025IDA-Bench: Evaluating LLMs on Interactive Guided Data Analysiscited by 11
  29. 2025AutoSDT: Scaling Data-Driven Discovery Tasks Toward Open Co-Scientistscited by 11
  30. 2025SAGE: An Agentic Explainer Framework for Interpreting SAE Features in Language Modelscited by 10
  31. 2025Evaluating Large Language Models as Expert Annotatorscited by 9
  32. 2026Human-AI Co-design for Clinical Prediction Modelscited by 9
  33. 2026HARBOR: Automated Harness Optimizationcited by 9
  34. 2026Test-time Recursive Thinking: Self-Improvement without External Feedbackcited by 7
  35. 2024Learning a Decision Tree Algorithm with Transformerscited by 6
  36. 2024"What is Different Between These Datasets?" A Framework for Explaining Data Distribution Shiftscited by 5
  37. 2023MaNtLE: Model-agnostic Natural Language Explainercited by 4
  38. 2026More Than "Means to an End": Supporting Reasoning with Transparently Designed AI Data Science Processescited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Can Language Model Agents be Helpful Circuit Explainers in Mechanistic Interpretability?cited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025Can Agentic AI Match the Performance of Human Data Scientists?
  2. 2026Pitfalls in Evaluating Interpretability Agents
  3. 2026Interpreting Agentic Systems: Beyond Model Explanations to System-Level Accountability
  4. 2026AgenticDataBench: A Comprehensive Benchmark for Data Agents
  5. 2026ADE: Agentic Data Evolution Framework for Human-Centered Objectives
  6. 2026Sanity Checks for Agentic Data Science
  7. 2026Vision: Data-Centric Anchoring for Robust and Interpretable Agentic AI
  8. 2026Deliberate Evolution: Agentic Reasoning for Sample-Efficient Symbolic Regression with LLMs