← 研踪 PaperTrailarXiv:2610.09471 · v5
When Should an In-Context Learner Expand Its Hypothesis Space?
首次提交 2026-10-07 13:26
版本更新 2026-11-08 22:48
雷达观测 2026-11-10 12:16
摘要
Learning systems adapt quickly within a familiar family of models. The harder step is deciding, from a prediction failure, whether opening a richer family is worth its cost. We cast this as a costly sequential decision and introduce the Structural Revision Environment (SRE), which produces matched failures from noise, exceptions, change within the family or structure outside it, varies the price of expansion and the remaining horizon independently of the evidence, and admits an exact normative solution of the one-shot decision. The solution shows that revision is a value boundary, not an evidence threshold: except at the lowest prices, the richer family becomes the more probable explanation long before expanding becomes optimal. Small Transformers trained in SRE reproduce the expansion boundary from utility alone. Pretrained 7B and 8B language models fail at a definite point: their predictions carry a failure-sensitive signal that their decisions do not reflect. Told the gain of expanding, they weigh it against price and horizon only when they reason, and the one model that finishes its reasoning on the history alone still does not estimate that gain. Controlled post-training of the small Transformers moves the prior and the sharpness of predictions, and neither moves the criterion.
作者备注：46 pages, 12 figures, 13 tables
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v5 · 2026-11-08 22:48 · 最新观测版本
When Should an In-Context Learner Expand Its Hypothesis Space?
Learning systems adapt quickly within a familiar family of models. The harder step is deciding, from a prediction failure, whether opening a richer family is worth its cost. We cast this as a costly sequential decision and introduce the Structural Revision Environment (SRE), which produces matched failures from noise, exceptions, change within the family or structure outside it, varies the price of expansion and the remaining horizon independently of the evidence, and admits an exact normative solution of the one-shot decision. The solution shows that revision is a value boundary, not an evidence threshold: except at the lowest prices, the richer family becomes the more probable explanation long before expanding becomes optimal. Small Transformers trained in SRE reproduce the expansion boundary from utility alone. Pretrained 7B and 8B language models fail at a definite point: their predictions carry a failure-sensitive signal that their decisions do not reflect. Told the gain of expanding, they weigh it against price and horizon only when they reason, and the one model that finishes its reasoning on the history alone still does not estimate that gain. Controlled post-training of the small Transformers moves the prior and the sharpness of predictions, and neither moves the criterion.
查看此版本解读 →查看此版本原文 ↗