Prune 'n Predict: Optimizing LLM Decision-making with Conformal Prediction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Vishwakarma, Harit, Mishler, Alan, Cook, Thomas, Dalmasso, Niccolò, Raman, Natraj, Ganesh, Sumitra
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916839507886080
author Vishwakarma, Harit
Mishler, Alan
Cook, Thomas
Dalmasso, Niccolò
Raman, Natraj
Ganesh, Sumitra
author_facet Vishwakarma, Harit
Mishler, Alan
Cook, Thomas
Dalmasso, Niccolò
Raman, Natraj
Ganesh, Sumitra
contents Large language models (LLMs) are empowering decision-making in several applications, including tool or API usage and answering multiple-choice questions (MCQs). However, incorrect outputs pose significant risks in high-stakes domains like healthcare and finance. To quantify LLM uncertainty and thereby mitigate these risks, recent works employ conformal prediction (CP), a model- and distribution-agnostic framework that uses LLM outputs to generate a \emph{prediction set} containing the true answer with high probability. Leveraging CP, we propose \emph{conformal revision of questions} (CROQ), which revises the question by narrowing down the available choices to those in the prediction set and asking the LLM the revised question. We expect LLMs to be more accurate on revised questions with fewer choices. Furthermore, we expect CROQ to be effective when the prediction sets from CP are small. Commonly used logit scores often lead to large sets, diminishing CROQ's effectiveness. To overcome this, we propose CP-OPT, an optimization framework to learn scores that minimize set sizes while maintaining coverage. Our extensive experiments on MMLU, ToolAlpaca, and TruthfulQA datasets with multiple LLMs show that CROQ improves accuracy over the standard inference, with more pronounced gains when paired with CP-OPT.
format Preprint
id arxiv_https___arxiv_org_abs_2501_00555
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Prune 'n Predict: Optimizing LLM Decision-making with Conformal Prediction
Vishwakarma, Harit
Mishler, Alan
Cook, Thomas
Dalmasso, Niccolò
Raman, Natraj
Ganesh, Sumitra
Machine Learning
Artificial Intelligence
Applications
Large language models (LLMs) are empowering decision-making in several applications, including tool or API usage and answering multiple-choice questions (MCQs). However, incorrect outputs pose significant risks in high-stakes domains like healthcare and finance. To quantify LLM uncertainty and thereby mitigate these risks, recent works employ conformal prediction (CP), a model- and distribution-agnostic framework that uses LLM outputs to generate a \emph{prediction set} containing the true answer with high probability. Leveraging CP, we propose \emph{conformal revision of questions} (CROQ), which revises the question by narrowing down the available choices to those in the prediction set and asking the LLM the revised question. We expect LLMs to be more accurate on revised questions with fewer choices. Furthermore, we expect CROQ to be effective when the prediction sets from CP are small. Commonly used logit scores often lead to large sets, diminishing CROQ's effectiveness. To overcome this, we propose CP-OPT, an optimization framework to learn scores that minimize set sizes while maintaining coverage. Our extensive experiments on MMLU, ToolAlpaca, and TruthfulQA datasets with multiple LLMs show that CROQ improves accuracy over the standard inference, with more pronounced gains when paired with CP-OPT.
title Prune 'n Predict: Optimizing LLM Decision-making with Conformal Prediction
topic Machine Learning
Artificial Intelligence
Applications
url https://arxiv.org/abs/2501.00555