跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.21393 · v1

Train After You Measure: a Pre-Registered Learning Curve for Fine-Tuning Typed Decision Models on One User's Labels

首次提交 2026-11-17 00:49
版本更新 2026-11-17 00:49
雷达观测 2026-11-17 16:10
arXiv ↗PDF ↗

摘要

Fine-tuning a small typed decision model on one user's email labels did not clear the margin we pre-registered over a logistic regression trained on the same labels. In earlier work, about two hundred labels and a TF-IDF logistic regression had the highest point accuracy on this inbox, though not significantly above the best hosted model; we ask whether the same labels, given to a decision model, close the gap. A first registered study, on a generic 4B base with cross-validation and message type only, stayed below the baseline at every label budget (-1.8 to -7.3 points). The main study was registered before any fine-tuning run: a fixed test set of 135 emails never used for selection, LoRA adapters on the public Tev1-0.8B and Tev1-4B decision models, nested subsets of 50 to 587 labels, two seeds and five decision fields on a free T4 GPU. Four of its five hypotheses are refuted--two in the fine-tunes' favour--and one is confirmed for one model. On message type the 4B model misses the registered three-point margin at 400 labels (+2.6) and passes it with all 587 (+3.3); with all labels both models are above the baseline by +0.7 and +3.3 points, where one email is 0.74 points and a 95% interval is about 15 points wide, and the 4B model is clearly ahead in type macro-F1 (0.581 vs. 0.406). On the three action switches the baseline is ahead at every label budget for both models (0.931 vs. 0.904 and 0.899 mean accuracy with all labels), at under a millisecond per question against 306 and 1681 ms. The 0.8B type curve barely improves after 300 labels and then falls. The baseline never predicts the two rarest classes (7 offers and 8 spam in the training pool); the fine-tuned models find spam in some runs, not consistently across seeds. We list the explanations we did not test and say when a new attempt would be worth making. Scripts are public; weights and emails are not.

作者备注:9 pages, 1 figure, 6 tables. Companion to arXiv:2610.32955. Pre-registered study; code at https://github.com/federicolucca/foyer-measure (tag v2.4; registrations in prereg/); weights and data not released

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

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。

v1 · 2026-11-17 00:49 · 最新观测版本

Train After You Measure: a Pre-Registered Learning Curve for Fine-Tuning Typed Decision Models on One User's Labels

Fine-tuning a small typed decision model on one user's email labels did not clear the margin we pre-registered over a logistic regression trained on the same labels. In earlier work, about two hundred labels and a TF-IDF logistic regression had the highest point accuracy on this inbox, though not significantly above the best hosted model; we ask whether the same labels, given to a decision model, close the gap. A first registered study, on a generic 4B base with cross-validation and message type only, stayed below the baseline at every label budget (-1.8 to -7.3 points). The main study was registered before any fine-tuning run: a fixed test set of 135 emails never used for selection, LoRA adapters on the public Tev1-0.8B and Tev1-4B decision models, nested subsets of 50 to 587 labels, two seeds and five decision fields on a free T4 GPU. Four of its five hypotheses are refuted--two in the fine-tunes' favour--and one is confirmed for one model. On message type the 4B model misses the registered three-point margin at 400 labels (+2.6) and passes it with all 587 (+3.3); with all labels both models are above the baseline by +0.7 and +3.3 points, where one email is 0.74 points and a 95% interval is about 15 points wide, and the 4B model is clearly ahead in type macro-F1 (0.581 vs. 0.406). On the three action switches the baseline is ahead at every label budget for both models (0.931 vs. 0.904 and 0.899 mean accuracy with all labels), at under a millisecond per question against 306 and 1681 ms. The 0.8B type curve barely improves after 300 labels and then falls. The baseline never predicts the two rarest classes (7 offers and 8 spam in the training pool); the fine-tuned models find spam in some runs, not consistently across seeds. We list the explanations we did not test and say when a new attempt would be worth making. Scripts are public; weights and emails are not.

查看此版本解读 →查看此版本原文 ↗