让大模型更准地从文本推断细微数值评分
Reasoning-Intensive Regression
- 用提示词优化+神经集成学习，提升文本数值推理精度
- 在四个真实任务上比基线提升最高达65%
- 适合数据少、需深度分析的场景，如评分、奖励建模
AI研究人员和从业者越来越多地将大语言模型（LLMs）应用于我们称之为推理密集型回归（RiR）的任务，即从文本中推断出微妙的数值评分。与情感分析或相似度评估等标准语言回归任务不同，RiR常见于临时性应用，如基于评分量表的评分、复杂环境中的密集奖励建模或特定领域检索，这类任务需要更深入的上下文分析，但可用的任务特定训练数据和计算资源极为有限。我们选取四个现实问题作为RiR任务，建立初始基准，并验证假设：仅通过提示冻结的LLM或通过梯度下降微调Transformer编码器，在RiR任务中通常表现不佳。为此，我们提出MENTAT——一种简单轻量的方法，结合批量反射式提示优化与神经集成学习。MENTAT在多个任务上相较两种基线最高提升65%，尽管仍有巨大改进空间。
原文摘要 · Abstract (English)
AI researchers and practitioners increasingly apply large language models (LLMs) to what we call reasoning-intensive regression (RiR), i.e., deducing subtle numerical scores from text. Unlike standard language regression tasks such as sentiment or similarity analysis, RiR often appears instead in ad-hoc applications such as rubric-based scoring, modeling dense rewards in complex environments, or domain-specific retrieval, where much deeper analysis of context is required while only limited task-specific training data and computation are available. We cast four realistic problems as RiR tasks to establish an initial benchmark, and use that to test our hypothesis that prompting frozen LLMs and fine-tuning Transformer encoders via gradient descent will both often struggle in RiR. We then propose MENTAT, a simple and lightweight method that combines batch-reflective prompt optimization with neural ensemble learning. MENTAT achieves up to 65% improvement over both baselines, though substantial room remains for future advances.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。