Metraipolis › archive › dc_42363ef64e0e
CEDAR-GRPO: Process-Aware Reinforcement Learning for General Abductive Reasoning in LLMs — arXiv:2608.14791
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.14791
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T17:08:06.982Z
- SHA-256 of the archived text
f176b3bb58082efaa6446d611475a20695fbed15618338bd73433f5f502b834d- Stable URL of this page
https://metraipolis.com/documento/dc_42363ef64e0e
How to cite
CEDAR-GRPO: Process-Aware Reinforcement Learning for General Abductive Reasoning in LLMs — arXiv:2608.14791 — arXiv (en), https://arxiv.org/abs/2608.14791 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_42363ef64e0e (Metraipolis archive dc_42363ef64e0e, sha256 f176b3bb58082efaa6446d611475a20695fbed15618338bd73433f5f502b834d)
BibTeX
@misc{dc42363ef64e0e,
title = {CEDAR-GRPO: Process-Aware Reinforcement Learning for General Abductive Reasoning in LLMs — arXiv:2608.14791},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.14791},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_42363ef64e0e; sha256 f176b3bb58082efaa6446d611475a20695fbed15618338bd73433f5f502b834d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.14791. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Abductive reasoning, often characterized as inference to the best explanation, is central to explanation under uncertainty, from everyday sense-making and investigation to scientific discovery. Yet LLM research has mostly studied abduction through narrow, task-specific benchmarks, making it unclear whether observed gains transfer beyond the benchmark family used for training or evaluation. We ask whether RL post-training can improve abduction as a transferable reasoning capability. We introduce CEDAR-GRPO, a process-aware framework that combines final-answer correctness with abductive rewards for evidence coverage and evidence-to-explanation directionality. Four open-weight LLMs are post-trained on a controlled, domain-neutral mixture of abductive hypothesis-generation and hypothesis-selection tasks. We evaluate them on 11 unseen tasks spanning hypothesis selection, missing-fact generation, defeasible inference, long-context investigation, clinical reasoning, code debugging, and non-abductive controls. CEDAR- GRPO improves every model on every held-out task over both base models and correctness-only GRPO, with average gains of 7.4 and 2.7 points, respectively, and a maximum gain of 30.8 points. Ablations confirm that RL, abductive reward design, and task diversity each contribute to transfer. Process-level metrics further show stronger abductive behavior, including exploration of alternatives, elimination of rivals, backtracking, and uncertainty marking. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.14791 — 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 2026-10-18T16:29:13.735Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).