通过增加推理时计算量，提升大模型问答自信度与准确率
Is That Your Final Answer? Test-Time Scaling Improves Selective Question Answering
- 推理时动态增加计算资源，提升模型判断能力
- 计算越多，正确回答的置信度越高，错误率下降18%
- 提出非零风险评估新范式，适合实际应用部署
大规模语言模型在推理时增加计算资源，显著提升了推理基准上的表现。然而，现有评估假设模型必须对所有问题给出答案，忽略了模型置信度与是否应答的问题。本文在推理过程中提取置信度分数，用于阈值筛选回答。结果表明，增加推理时计算预算不仅使模型更准确地回答问题，还提高了正确回答的置信度。我们进一步拓展了零风险响应的评估范式，引入非零响应风险场景，并提出相应的评估报告方法。
原文摘要 · Abstract (English)
Scaling the test-time compute of large language models has demonstrated impressive performance on reasoning benchmarks. However, existing evaluations of test-time scaling make the strong assumption that a reasoning system should always give an answer to any question provided. This overlooks concerns about whether a model is confident in its answer, and whether it is appropriate to always provide a response. To address these concerns, we extract confidence scores during reasoning for thresholding model responses. We find that increasing compute budget at inference time not only helps models answer more questions correctly, but also increases confidence in correct responses. We then extend the current paradigm of zero-risk responses during evaluation by considering settings with non-zero levels of response risk, and suggest a recipe for reporting evaluations under these settings.
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。