The paper's descriptive claim that beliefs affect RLHF preferences is supported by a priming study, but its main normative bound (Theorem 4.3) is not actually derived and appears false as stated.
Carla: An open urban driving simulator
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
A Descriptive and Normative Theory of Human Beliefs in RLHF
The paper's descriptive claim that beliefs affect RLHF preferences is supported by a priming study, but its main normative bound (Theorem 4.3) is not actually derived and appears false as stated.