Contrastive Discovery 提出对比集方法，让 LLM 科学发现智能体学会选择替代解释
Contrastive Discovery: Open-Ended Scientific Discovery over Competing Explanations
做科学发现智能体的开发者该看看这篇：智能体跑完实验只会说"假设被拒绝了"，却答不出哪个解释成立，论文用对比集把 73% 的拒绝转成了明确答案，同等预算下多解决 67% 的问题。改动只在假设的提出单元上，现有智能体框架可以直接套用。
一篇 arXiv 论文（编号 2611.14514）提出 Contrastive Discovery 框架，针对现有 LLM 科学发现智能体只做逐点假设检验的缺陷：当证据拒绝假设 h 时，智能体只知道 h 不成立，却不知道哪个替代解释成立。
核心机制是把单个假设替换为"对比集"——一个科学问题配上一组互斥且穷尽竞争解释。每个实验按其对集合内各解释信念重分配的程度打分，当信念集中到某个解释上即视为问题解决。由于只改变了智能体提出和评分的单元，现有智能体可直接用对比集复评已收集的证据或引导搜索。
实验结果：跨六个领域，逐点方法下信念偏移最大的候选中 89% 是拒绝；同一批证据下，对比集将其中 73% 解析到具体替代解释。用对比集引导搜索时，同等实验预算下解析的问题数比逐点 surprisal 方法多 67%。在 38 位人类专家参与的盲测中，专家在 78% 的已解析问题上选出了与智能体相同的解释。
Contrastive Discovery: Open-Ended Scientific Discovery over Competing Explanations
LLM agents are increasingly used for open-ended, data-driven scientific discovery, yet most of them evaluate hypotheses pointwise, testing an empirical claim $h$ only against its complement $\neg h$. When evidence rejects $h$, the agent learns that one explanation is unlikely but not which alternative the evidence supports. To recover that alternative, we introduce Contrastive Discovery, which replaces the single hypothesis with a contrast set, a scientific question paired with mutually exclusive and collectively exhaustive competing explanations. Each experiment is scored by how its evidence redistributes belief across the set, and the question is resolved when belief concentrates on one explanation. Because contrast sets change only the unit an agent proposes and scores, an existing agent can use them to re-evaluate evidence it has already collected or to guide its search. Across six domains, $89\%$ of the pointwise candidates with the largest belief shifts are rejections. On the same evidence, contrast sets resolve $73\%$ of these candidates to a specific alternative. When contrast sets guide search, the agent resolves $67\%$ more questions than pointwise surprisal under the same experiment budget. In a blinded study, $38$ human experts chose the same explanation on $78\%$ of resolved questions and on $82\%$ of questions where the pointwise agent rejected its hypothesis and stopped. Together, these results support explicit competing explanations as a way for discovery agents to resolve more scientific questions from the same evidence and validation budget.
本条情报归属 AITOP「论文」核心栏目，重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：LLM智能体、科学发现、arXiv论文、假设检验，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.AI」（@Ziang Liu, James J. Kim, Yijia Dai, Jennifer J. Sun），经 AITOP 多维评分管线结构化提炼。
