Metraipolis › archive › dc_c70539b968c5
J-Miner: Recovering the Decision Logic of Fine-Tuned LLM Classifiers as Compact Rules — arXiv:2608.17063
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.17063
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T14:48:26.685Z
- SHA-256 of the archived text
fb3abd06478a94ee4bb292de15dee67fa2cdf1becd6008ffdccbc28d8e958d26- Stable URL of this page
https://metraipolis.com/documento/dc_c70539b968c5
How to cite
J-Miner: Recovering the Decision Logic of Fine-Tuned LLM Classifiers as Compact Rules — arXiv:2608.17063 — arXiv (en), https://arxiv.org/abs/2608.17063 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c70539b968c5 (Metraipolis archive dc_c70539b968c5, sha256 fb3abd06478a94ee4bb292de15dee67fa2cdf1becd6008ffdccbc28d8e958d26)
BibTeX
@misc{dcc70539b968c5,
title = {J-Miner: Recovering the Decision Logic of Fine-Tuned LLM Classifiers as Compact Rules — arXiv:2608.17063},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.17063},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c70539b968c5; sha256 fb3abd06478a94ee4bb292de15dee67fa2cdf1becd6008ffdccbc28d8e958d26},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.17063. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Task-fine-tuned large language model (LLM) classifiers acquire task-specific decision knowledge, but this knowledge remains implicit in distributed internal computations, making their decision logic difficult to interpret. We introduce the Executable Decision Compression (EDC) framework and propose J-Miner, which mines vocabulary-named variables from internal readouts and learns rules shared across inputs to produce executable explanations. Analysis reveals that a small set of these variables captures much of the classifier's decision behavior, holding for both varying parameter scales within a family and distinct families. Across six binary tasks, a rule using just one variable reproduces 76.7% of source-classifier decisions on average, rising to 88.8% with 16 variables. Most of the decision information retained by these variables comes from internal activations beyond literal surface matching. A lightweight text reader predicts the variable states, allowing the same fixed rules to execute independently of the source classifier. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-19 in cs.LG). Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.17063 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T14:47:59.375Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-19?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).