arXiv cs.CL· Shani Goren, Ido Galil, Ran El-Yaniv·· 12 小时前AI 评分38
论文提出 Selective Abstraction：让 LLM 对不确定内容降低具体程度以提升长文可靠性
When Should LLMs Be Less Specific? Selective Abstraction for Reliable Long-Form Text Generation
AI 导读
研究者提出 Selective Abstraction（SA）框架，让 LLM 通过选择性降低不确定内容的细节程度，在具体性与可靠性之间做权衡。
正文 · 原文
Abstract:LLMs are widely used, yet they remain prone to factual errors that erode trust and limit adoption in high-risk settings. One approach to mitigate this risk is to equip models with uncertainty estimation mechanisms that abstain when confidence is low. However, this binary "all-or-nothing" approach is excessively restrictive in long-form settings, often discarding valuable information. We introduce Selective Abstraction (SA), a framework that enables LLMs to trade specificity for reliability by selectively reducing the detail of uncertain content. We formalize SA through selective risk and coverage, and propose Atom-wise Selective Abstraction, a claim-level instantiation that decomposes responses into atomic claims and replaces uncertain atoms with higher-confidence, less specific abstractions. We develop a novel end-to-end pipeline for evaluating long-form generation that instantiates risk as factual correctness and measures coverage as retained information. Our comprehensive empirical evaluation across five LLMs and two long-form factual benchmarks shows that atom-wise SA consistently improves the risk-coverage trade-off, outperforming seven baselines, including claim removal, self-revision, sampling-based aggregation, and chain-of-verification. It improves AURC by up to 35.61% over claim removal, demonstrating that reducing specificity boosts reliability while preserving substantial information. We further provide a risk-guided procedure for selecting a confidence threshold that satisfies a target risk.
| Comments: | Accepted as poster to NeurIPS 2026 |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) |
| Cite as: | arXiv:2602.11908 [cs.AI] |
| (or arXiv:2602.11908v4 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2602.11908 arXiv-issued DOI via DataCite |
Submission history
From: Shani Goren [view email]
[v1]
Thu, 12 Feb 2026 13:06:14 UTC (640 KB)
[v2]
Fri, 13 Feb 2026 13:48:01 UTC (640 KB)
[v3]
Mon, 1 Jun 2026 21:45:36 UTC (640 KB)
[v4]
Sun, 1 Nov 2026 11:42:50 UTC (741 KB)
来源：arXiv cs.CL · arxiv.org