Date: 25/06/2025 21:37:20
From: dv
ID: 2295634
Subject: re: Reward Model Interpretability via Optimal and Pessimal Tokens

esselte said:


The Rev Dodgson said:

dv said:

https://dl.acm.org/doi/10.1145/3715275.3732068

Reward Model Interpretability via Optimal and Pessimal Tokens
Authors: Brian Christian, Hannah Rose Kirk, Jessica A.F. Thompson, Christopher Summerfield, Tsvetomira Dumbalska
FAccT ’25: Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency


“Reward modeling has emerged as a crucial component in aligning large language models with human values.
Significant attention has focused on using reward models as a means for fine
tuning generative models. However, the reward models themselves— which directly encode human value judgments by turning prompt
response pairs into scalar rewards—remain relatively understudied. We present a novel approach to reward model interpretability through exhaustive analysis of their responses across their entire vocabulary space.”

I sort of get what they mean, but how that relates to the chart is a mystery to me.

dv is a “memelord”… He’s taken the least interesting but most memey aspect of this paper and presented it as it’s own thread.

I was hoping at least for Ataturk

Reply Quote View full thread