Metraipolis › archive › dc_a571b24311e2
CoEvo: Oracle-Grounded Self-Evolution of a Single Model for Multi-Step Causal Reasoning — arXiv:2609.26094
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.26094
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T15:07:46.018Z
- SHA-256 of the archived text
96a31dcd88d716eeb9c8a430f0e89e74e78e58b8eb216bd5748b1b2219f27a3d- Stable URL of this page
https://metraipolis.com/documento/dc_a571b24311e2
How to cite
CoEvo: Oracle-Grounded Self-Evolution of a Single Model for Multi-Step Causal Reasoning — arXiv:2609.26094 — arXiv (en), https://arxiv.org/abs/2609.26094 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a571b24311e2 (Metraipolis archive dc_a571b24311e2, sha256 96a31dcd88d716eeb9c8a430f0e89e74e78e58b8eb216bd5748b1b2219f27a3d)
BibTeX
@misc{dca571b24311e2,
title = {CoEvo: Oracle-Grounded Self-Evolution of a Single Model for Multi-Step Causal Reasoning — arXiv:2609.26094},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.26094},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a571b24311e2; sha256 96a31dcd88d716eeb9c8a430f0e89e74e78e58b8eb216bd5748b1b2219f27a3d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.26094. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-step causal reasoning requires chaining inferences where each step constrains the next. An early error propagates silently, and a correct answer reached via flawed logic evades outcome-level detection. In specialized domains, teacher LLMs err on intermediate steps, safety constraints restrict cloud distillation, and shifting conditions demand adaptation, leaving self-evolution as the practical route. Naive self-evolution can collapse: outcome-only rewards let the model exploit distributional shortcuts, and weak self-evaluation reinforces spurious paths into stable failure patterns. We exploit a key asymmetry: generating a correct chain is hard, but verifying a single step is easy. Many high-stakes domains admit a deterministic, queryable oracle, a physics simulator or rule engine over codified constraints. It checks asserted steps without teacher-level ability and abstains beyond its rules; it can check what the model asserts, never replace it. This enables CoEvo, an oracle-grounded self-evolution framework where a single model alternates between Proposer and Solver. As Solver, the model generates competing chains; intra-group debate exposes disagreement steps, a proxy for the capability boundary, and the oracle adjudicates them into process-level supervision. As Proposer, the same model constructs progressively harder scenarios inside oracle constraints, steering the curriculum toward deep multi-hop chains. Both roles are updated jointly, so training pressure co-evolves with the model. On industrial, clinical, and legal multi-step causal reasoning benchmarks, CoEvo enables an 8B LLM to sustain self-evolution, surpassing distillation baselines and the strongest proprietary reference on path correctness (82.1% vs. 71.4%). The trained model generalizes to unseen categories and systems, preserving root-cause accuracy. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.26094 — 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 2026-10-08T14:37:29.457Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).