Goodhart in the Planner：基于采样的控制会利用从人类偏好中学到的奖励

View PDF HTML (experimental)

Abstract:Robots increasingly plan by searching against a reward model learned from human preferences; we measure what this search does to the true objective. On HalfCheetah and Hopper with crowd-sourced labels, and on Meta-World drawer-open and plate-slide with synthetic labels and blind labels from the two authors, harder MPPI search cuts true return by up to 76% while held-out label accuracy stays at 0.75-0.97: with human labels the drawer opens in 0 of 20 runs, against 10 of 10 for the cloned policy the planner starts from. The effect persists with a learned dynamics model and with an image scorer. True return roughly follows how far the planner's weights move from its proposals ($\sqrt{\mathrm{KL}}$), and best-of-$N$ action selection with a verifier trained on synthetic labels falls below a single noisy action. We prove that $\chi^2$-regularized MPPI at a temperature set by the scorer's error spread on the proposals is worst-case optimal, that no gain can be certified once that spread exceeds the scorer's own, and that an episode's return splits exactly into per-step terms, only some of which the guarantee covers. Calibrated from 100 proposal pairs ranked by true return (an oracle), this rule is the safest fix we test: it keeps on average 86-94% of the cloned policy's return on drawer-open and plate-slide, though it is cautious on window-open, where search helps. Calibrated from a person's own comparisons, it cannot detect a bias the scorer shares with that person; a second person, or correcting for pairs people cannot tell apart, limits the damage.

Submission history

From: Soovadeep Bakshi [view email]
[v1] Fri, 6 Nov 2026 16:22:17 UTC (514 KB)
[v2] Thu, 12 Nov 2026 07:40:41 UTC (517 KB)