Metraipolis › archive › dc_2e4897ea13ed
Auditability Is Not One Property: Rule Overlap, Behavioural Agreement, and Composition in Reinforcement Learning — arXiv:2609.28581
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.28581
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T04:26:55.918Z
- SHA-256 of the archived text
1e5ee13daa05fbb7f0476654b5895f92569e552befe470322aa64bf5e716f3bb- Stable URL of this page
https://metraipolis.com/documento/dc_2e4897ea13ed
How to cite
Auditability Is Not One Property: Rule Overlap, Behavioural Agreement, and Composition in Reinforcement Learning — arXiv:2609.28581 — arXiv (en), https://arxiv.org/abs/2609.28581 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2e4897ea13ed (Metraipolis archive dc_2e4897ea13ed, sha256 1e5ee13daa05fbb7f0476654b5895f92569e552befe470322aa64bf5e716f3bb)
BibTeX
@misc{dc2e4897ea13ed,
title = {Auditability Is Not One Property: Rule Overlap, Behavioural Agreement, and Composition in Reinforcement Learning — arXiv:2609.28581},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.28581},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2e4897ea13ed; sha256 1e5ee13daa05fbb7f0476654b5895f92569e552befe470322aa64bf5e716f3bb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28581. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Reinforcement-learning (RL) policies are often distributed as opaque neural checkpoints, while training logs show that a run occurred without explaining what the policy learned. We study whether independently trained policies can be represented and composed through auditable discrete behavioral rules. We define auditability as six separately testable predicates: trace integrity, lossless coding, rule coverage, behavioral agreement, composition quality, and value-model reliability. Our protocol uses a shared frozen symbolizer, passive rule extraction, an append-only hash-bound ledger, exact environment replay, and offline confidence-ranked arbitration with an explicit blind-spot fallback. The results place strict limits on this description layer. Rule-set overlap does not imply behavioral agreement: policies may share symbolic rules while choosing near-chance-matching actions on fresh states. The fused policy therefore selects among existing rules rather than generating a new skill. On a conflict-dominated task, an apparent fusion failure is traced to an induction/deployment mismatch: rules induced from sampled actions were evaluated under argmax actions, and deployment-consistent re-induction reverses the arbitration ordering. A fitted-Q generalized-policy-improvement diagnostic also fails in both environments, limiting claims that rule fusion is superior to value-based composition. One exploratory comparison favors rule fusion, but its comparator is post hoc, the task is partly saturated, and the fused policy remains below the strongest held-out actor. We contribute an evidence-bounded audit and composition protocol, not a claim of universal interpretability or autonomous skill generation. Future work must add temporally extended skills, cross-skill interfaces, composition search, and independent novelty audits. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.28581 — 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 2026-10-08T03:50:57.926Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).