[2412.00543] Evaluating the Consistency of LLM Evaluators
Abstract:Large language models (LLMs) have shown potential as general evaluators along with the evident benefits of speed and cost. While their correlation against human annotators has been widely studied, consistency as evaluators is still understudied, raising concerns about the reliability of LLM evaluators. In this paper, we conduct extensive studies on the two aspects of consistency in LLM evaluations, Self-Consistency (SC) and Inter-scale Consistency (IC), on different scoring scales and criterion granularity with open-source and proprietary models. Our comprehensive analysis demonstrates that strong proprietary models are not necessarily consistent evaluators, highlighting the importance of considering consistency in assessing the capability of LLM evaluators.
Evaluating the Consistency of LLM Evaluators Noah Lee∗ Jiwoo Hong∗ James Thorne KAIST AI {noah.lee, jiwoo_hong, thorne}@kaist.ac.kr Abstract 2024; Li et al., 2024). However, the consistency of…
saved by
related reading
- Applying Statistics to LLM Evaluationscameronrwolfe.substack.com
- The bitter lesson of LLM evalsparsed.com
- [2411.00640] Adding Error Bars to Evals: A Statistical Approach to Language Model Evaluationsarxiv.org
- Your AI Product Needs Evals – Hamel's Blog - Hamel Husainhamel.dev
- LLM Evaluation doesn't need to be complicatedphilschmid.de
- What We’ve Learned From A Year of Building with LLMs – Applied LLMsapplied-llms.org
- Successful language model evals - Jason Weijasonwei.net
- LLM-as-a-judge: a complete guide to using LLMs for evaluationsevidentlyai.com
- Patterns for Building LLM-based Systems & Productseugeneyan.com
- Position: It's Time to Optimize for Self-Consistencytime-for-consistency.github.io
- Discovering Language Model Behaviors with Model-Written Evaluations — LessWronglesswrong.com
- How to Use LLM as a Judge (Without Getting Burned)manthanguptaa.in