Saved in:
Bibliographic Details
Main Authors: Costarelli, Anthony, Allen, Mat, Field, Severin
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.02472
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • As Large Language Models (LLMs) become increasingly integrated into our daily lives, the potential harms from deceptive behavior underlie the need for faithfully interpreting their decision-making. While traditional probing methods have shown some effectiveness, they remain best for narrowly scoped tasks while more comprehensive explanations are still necessary. To this end, we investigate meta-models-an architecture using a "meta-model" that takes activations from an "input-model" and answers natural language questions about the input-model's behaviors. We evaluate the meta-model's ability to generalize by training them on selected task types and assessing their out-of-distribution performance in deceptive scenarios. Our findings show that meta-models generalize well to out-of-distribution tasks and point towards opportunities for future research in this area. Our code is available at https://github.com/acostarelli/meta-models-public .