InnerThoughts: Disentangling Representations and Predictions in Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chételat, Didier, Cotnareanu, Joseph, Thompson, Rylee, Zhang, Yingxue, Coates, Mark
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929692175499264
author Chételat, Didier
Cotnareanu, Joseph
Thompson, Rylee
Zhang, Yingxue
Coates, Mark
author_facet Chételat, Didier
Cotnareanu, Joseph
Thompson, Rylee
Zhang, Yingxue
Coates, Mark
contents Large language models (LLMs) contain substantial factual knowledge which is commonly elicited by multiple-choice question-answering prompts. Internally, such models process the prompt through multiple transformer layers, building varying representations of the problem within its hidden states. Ultimately, however, only the hidden state corresponding to the final layer and token position are used to predict the answer label. In this work, we propose instead to learn a small separate neural network predictor module on a collection of training questions, that take the hidden states from all the layers at the last temporal position as input and outputs predictions. In effect, such a framework disentangles the representational abilities of LLMs from their predictive abilities. On a collection of hard benchmarks, our method achieves considerable improvements in performance, sometimes comparable to supervised fine-tuning procedures, but at a fraction of the computational cost.
format Preprint
id arxiv_https___arxiv_org_abs_2501_17994
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle InnerThoughts: Disentangling Representations and Predictions in Large Language Models
Chételat, Didier
Cotnareanu, Joseph
Thompson, Rylee
Zhang, Yingxue
Coates, Mark
Computation and Language
Machine Learning
Large language models (LLMs) contain substantial factual knowledge which is commonly elicited by multiple-choice question-answering prompts. Internally, such models process the prompt through multiple transformer layers, building varying representations of the problem within its hidden states. Ultimately, however, only the hidden state corresponding to the final layer and token position are used to predict the answer label. In this work, we propose instead to learn a small separate neural network predictor module on a collection of training questions, that take the hidden states from all the layers at the last temporal position as input and outputs predictions. In effect, such a framework disentangles the representational abilities of LLMs from their predictive abilities. On a collection of hard benchmarks, our method achieves considerable improvements in performance, sometimes comparable to supervised fine-tuning procedures, but at a fraction of the computational cost.
title InnerThoughts: Disentangling Representations and Predictions in Large Language Models
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2501.17994