Dipper: Diversity in Prompts for Producing Large Language Model Ensembles in Reasoning tasks

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Lau, Gregory Kang Ruey, Hu, Wenyang, Liu, Diwen, Chen, Jizhuo, Ng, See-Kiong, Low, Bryan Kian Hsiang
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918170438139904
author Lau, Gregory Kang Ruey
Hu, Wenyang
Liu, Diwen
Chen, Jizhuo
Ng, See-Kiong
Low, Bryan Kian Hsiang
author_facet Lau, Gregory Kang Ruey
Hu, Wenyang
Liu, Diwen
Chen, Jizhuo
Ng, See-Kiong
Low, Bryan Kian Hsiang
contents Large Language Models (LLMs), particularly smaller variants, still struggle with complex reasoning tasks. While inference-time prompting can guide reasoning, existing methods often rely on sequential queries. Ensemble approaches offer a promising path to performance gains, especially given recent batch inference speed-ups. This work introduces DIPPER, a novel, training-free framework that transforms a single LLM into an effective inference-time ensemble. By feeding the model an optimized and diverse set of prompts in parallel, DIPPER elicits varied reasoning paths, leading to performance gains. We empirically demonstrate significant improvements on reasoning benchmarks, such as MATH, where a DIPPER ensemble of three Qwen2-MATH-1.5B instances (via parallel prompting of a single model) outperforms a larger 7B model.
format Preprint
id arxiv_https___arxiv_org_abs_2412_15238
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Dipper: Diversity in Prompts for Producing Large Language Model Ensembles in Reasoning tasks
Lau, Gregory Kang Ruey
Hu, Wenyang
Liu, Diwen
Chen, Jizhuo
Ng, See-Kiong
Low, Bryan Kian Hsiang
Computation and Language
Artificial Intelligence
Machine Learning
Multiagent Systems
Large Language Models (LLMs), particularly smaller variants, still struggle with complex reasoning tasks. While inference-time prompting can guide reasoning, existing methods often rely on sequential queries. Ensemble approaches offer a promising path to performance gains, especially given recent batch inference speed-ups. This work introduces DIPPER, a novel, training-free framework that transforms a single LLM into an effective inference-time ensemble. By feeding the model an optimized and diverse set of prompts in parallel, DIPPER elicits varied reasoning paths, leading to performance gains. We empirically demonstrate significant improvements on reasoning benchmarks, such as MATH, where a DIPPER ensemble of three Qwen2-MATH-1.5B instances (via parallel prompting of a single model) outperforms a larger 7B model.
title Dipper: Diversity in Prompts for Producing Large Language Model Ensembles in Reasoning tasks
topic Computation and Language
Artificial Intelligence
Machine Learning
Multiagent Systems
url https://arxiv.org/abs/2412.15238