Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Rajamohan, Srijith, Salhin, Ahmed, Frazier, Josh, Kumar, Rohit, Tsai, Yu-Cheng, Cook, Todd
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2502.08631
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917928479227904
author Rajamohan, Srijith
Salhin, Ahmed
Frazier, Josh
Kumar, Rohit
Tsai, Yu-Cheng
Cook, Todd
author_facet Rajamohan, Srijith
Salhin, Ahmed
Frazier, Josh
Kumar, Rohit
Tsai, Yu-Cheng
Cook, Todd
contents The output of Large Language Models (LLMs) are a function of the internal model's parameters and the input provided into the context window. The hypothesis presented here is that under a greedy sampling strategy the variance in the LLM's output is a function of the conceptual certainty embedded in the model's parametric knowledge, as well as the lexical variance in the input. Finetuning the model results in reducing the sensitivity of the model output to the lexical input variations. This is then applied to a classification problem and a probabilistic method is proposed for estimating the certainties of the predicted classes.
format Preprint
id arxiv_https___arxiv_org_abs_2502_08631
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Ensemble based approach to quantifying uncertainty of LLM based classifications
Rajamohan, Srijith
Salhin, Ahmed
Frazier, Josh
Kumar, Rohit
Tsai, Yu-Cheng
Cook, Todd
Artificial Intelligence
The output of Large Language Models (LLMs) are a function of the internal model's parameters and the input provided into the context window. The hypothesis presented here is that under a greedy sampling strategy the variance in the LLM's output is a function of the conceptual certainty embedded in the model's parametric knowledge, as well as the lexical variance in the input. Finetuning the model results in reducing the sensitivity of the model output to the lexical input variations. This is then applied to a classification problem and a probabilistic method is proposed for estimating the certainties of the predicted classes.
title Ensemble based approach to quantifying uncertainty of LLM based classifications
topic Artificial Intelligence
url https://arxiv.org/abs/2502.08631