Metraipolis › archive › dc_eb79d4b3cf88
Why2Speak: Faithful Reasoning for Abstaining Action Policies — arXiv:2608.20670
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.20670
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T10:18:38.299Z
- SHA-256 of the archived text
0f4f9b6c89aa9e35323f2ad726d9ef3b2dc921f0c0a07fb6720cc236a24b0ec1- Stable URL of this page
https://metraipolis.com/documento/dc_eb79d4b3cf88
How to cite
Why2Speak: Faithful Reasoning for Abstaining Action Policies — arXiv:2608.20670 — arXiv (en), https://arxiv.org/abs/2608.20670 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eb79d4b3cf88 (Metraipolis archive dc_eb79d4b3cf88, sha256 0f4f9b6c89aa9e35323f2ad726d9ef3b2dc921f0c0a07fb6720cc236a24b0ec1)
BibTeX
@misc{dceb79d4b3cf88,
title = {Why2Speak: Faithful Reasoning for Abstaining Action Policies — arXiv:2608.20670},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.20670},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eb79d4b3cf88; sha256 0f4f9b6c89aa9e35323f2ad726d9ef3b2dc921f0c0a07fb6720cc236a24b0ec1},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.20670. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Many agentic systems must repeatedly choose between acting and abstaining, making faithful reasoning important for oversight: an explanation is useful only if it reflects the computation that produced the action. We study this problem through intervention timing in multi-party conversation, where an assistant must decide whether to speak or remain silent. This setting exposes class imbalance, asymmetric action costs, and the possibility that exposing reasoning changes the policy being audited. Using Qwen3-8B, decoded with or without chain-of-thought reasoning, we compare direct decision policies, reasoning policies, supervised fine-tuning, and reinforcement learning. We find a capability-auditability tradeoff: the strongest direct policy achieves higher quality but exposes no reasoning to inspect, while the reasoning policy provides a trace at the cost of lower performance, particularly recall of true intervention opportunities. Supervised fine-tuning either suppresses reasoning or preserves it without improving decision quality, while reinforcement learning also fails to improve the reasoning policy. We identify one mechanism underlying this failure: group relative objectives provide no learning signal on confidently wrong prompts when sampled rollouts all select the same action. Controlled activation probes and behavioral ablations show that standard faithfulness methods can overstate evidence that exposed reasoning reflects the underlying decision process. Probability-based metrics saturate under confident decisions, probes are vulnerable to class imbalance and textual leakage, and reasoning ablations can confound reasoning content with changes in inference mode. Together, these results show that exposing reasoning can change an agent's action policy rather than simply make it observable. We provide controls for evaluating reasoning-based oversight of agents that can act or abstain. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-24 in cs.CL). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.20670 — 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 2026-10-18T10:17:37.611Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).