analiza aiqbench
This arXiv paper uses LLM coding agents to iteratively write interpretable rule sets for classification, reporting competitive results on small tabular datasets and larger gains on text. The useful takeaway for AI readers is a new framing: coding agents can serve as a flexible program-search engine for interpretable models, and their performance is measured against established rule-learning methods on test-set error. Claims rest on benchmark comparisons from an early preprint, so robustness across datasets and the cost of LLM-driven search remain open questions.
Coding LLMs Can Be Harnessed to Learn Interpretable Models from Data
arXiv:2610.14006v1 Announce Type: cross Abstract: Since compact, interpretable classifiers correspond to simple programs, it is natural to ask if LLM-based coding agents can be used to learn interpretable models from data. In this paper we present a simple harness that repeatedly prompts a coding LLM to produce the next rule in a rule set, given a sample of labeled examples. We show that this approach is competitive with, and often superior to, state-of-the-art rule-learning methods on tabular data, especially with small training sets, and wins by large margins on text, including against another LLM-based rule learner; that this approach outperforms standard harnesses, and that the approach immediately leads to new kinds of interpretable models, since coding agents are not restricted to search a highly restricted hypothesis space. In addition to the advances made in learning interpretable models, these results are of interest because they introduce a new class of coding benchmarks - construction of interpretable classifiers - which includes quantitative measurements of output quality (i.e., test-set error) and existing mature ``specialist coding algorithms'' (i.e., existing learning methods).
Abstract:Since compact, interpretable classifiers correspond to simple programs, it is natural to ask if LLM-based coding agents can be used to learn interpretable models from data. In this paper we present a simple harness that repeatedly prompts a coding LLM to produce the next rule in a rule set, given a sample of labeled examples. We show that this approach is competitive with, and often superior to, state-of-the-art rule-learning methods on tabular data, especially with small training sets, and wins by large margins on text, including against another LLM-based rule learner; that this approach outperforms standard harnesses, and that the approach immediately leads to new kinds of interpretable models, since coding agents are not restricted to search a highly restricted hypothesis space. In addition to the advances made in learning interpretable models, these results are of interest because they introduce a new class of coding benchmarks - construction of interpretable classifiers - which includes quantitative measurements of output quality (i.e., test-set error) and existing mature ``specialist coding algorithms'' (i.e., existing learning methods).
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2610.14006 [cs.LG]
(or arXiv:2610.14006v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.14006 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 269 cuvinte
Subiecte conexe
NVIDIA GTC Taipei at COMPUTEX: Live Updates on What’s Next in AI
NVIDIA · 21 mai 2026
[AINews] Here are 6 Clones of Jev in 2 days
Latent Space · 19 sept. 2026
Humanity’s Last Invention — Richard Socher of Recursive
Latent Space · 14 sept. 2026
The Knowledge Layer for Enterprise AI
Neo4j · 28 sept. 2026