Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wang, Wanying, Ma, Zeyu, Wang, Xuhong, Zhang, Yangchun, Liu, Pengfei, Chen, Mingang
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2410.11507
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908557220249600
author Wang, Wanying
Ma, Zeyu
Wang, Xuhong
Zhang, Yangchun
Liu, Pengfei
Chen, Mingang
author_facet Wang, Wanying
Ma, Zeyu
Wang, Xuhong
Zhang, Yangchun
Liu, Pengfei
Chen, Mingang
contents As Large Language Models (LLMs) are increasingly deployed in highly specialized vertical domains, the evaluation of their domain-specific performance becomes critical. However, existing evaluations for vertical domains typically rely on the labor-intensive construction of static single-turn datasets, which present two key limitations: (i) manual data construction is costly and must be repeated for each new domain, and (ii) static single-turn evaluations are misaligned with the dynamic multi-turn interactions in real-world applications, limiting the assessment of professionalism and stability. To address these, we propose TestAgent, a framework for automatic benchmarking and exploratory dynamic evaluation in vertical domains. TestAgent leverages retrieval-augmented generation to create domain-specific questions from user-provided knowledge sources, combined with a two-stage criteria generation process, thereby enabling scalable and automated benchmark creation. Furthermore, it introduces a reinforcement learning-guided multi-turn interaction strategy that adaptively determines question types based on real-time model responses, dynamically probing knowledge boundaries and stability. Extensive experiments across medical, legal, and governmental domains demonstrate that TestAgent enables efficient cross-domain benchmark generation and yields deeper insights into model behavior through dynamic exploratory evaluation. This work establishes a new paradigm for automated and in-depth evaluation of LLMs in vertical domains.
format Preprint
id arxiv_https___arxiv_org_abs_2410_11507
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TestAgent: Automatic Benchmarking and Exploratory Interaction for Evaluating LLMs in Vertical Domains
Wang, Wanying
Ma, Zeyu
Wang, Xuhong
Zhang, Yangchun
Liu, Pengfei
Chen, Mingang
Artificial Intelligence
Computation and Language
As Large Language Models (LLMs) are increasingly deployed in highly specialized vertical domains, the evaluation of their domain-specific performance becomes critical. However, existing evaluations for vertical domains typically rely on the labor-intensive construction of static single-turn datasets, which present two key limitations: (i) manual data construction is costly and must be repeated for each new domain, and (ii) static single-turn evaluations are misaligned with the dynamic multi-turn interactions in real-world applications, limiting the assessment of professionalism and stability. To address these, we propose TestAgent, a framework for automatic benchmarking and exploratory dynamic evaluation in vertical domains. TestAgent leverages retrieval-augmented generation to create domain-specific questions from user-provided knowledge sources, combined with a two-stage criteria generation process, thereby enabling scalable and automated benchmark creation. Furthermore, it introduces a reinforcement learning-guided multi-turn interaction strategy that adaptively determines question types based on real-time model responses, dynamically probing knowledge boundaries and stability. Extensive experiments across medical, legal, and governmental domains demonstrate that TestAgent enables efficient cross-domain benchmark generation and yields deeper insights into model behavior through dynamic exploratory evaluation. This work establishes a new paradigm for automated and in-depth evaluation of LLMs in vertical domains.
title TestAgent: Automatic Benchmarking and Exploratory Interaction for Evaluating LLMs in Vertical Domains
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2410.11507