Evaluating Large Language Models at Evaluating Instruction Following
Preprint
Abstract
As research in large language models (LLMs) continues to accelerate, LLM-based evaluation has emerged as a scalable and cost-effective alternative to human evaluations for comparing the ever increasing list of models. This paper investigates the efficacy of these ``LLM evaluators'', particularly in using them to assess instruction following, a metric that gauges how closely generated text adheres to the given instruction. We introduce a challenging meta-evaluation benchmark, LLMBar, designed to test the ability of an LLM evaluator in discerning instruction-following outputs. The authors manually curated 419 pairs of outputs, one adhering to instructions while the other diverging, yet may possess deceptive qualities that mislead an LLM evaluator, e.g., a more engaging tone. Contrary to existing meta-evaluation, we discover that different evaluators (i.e., combinations of LLMs and prompts) exhibit distinct performance on LLMBar and even the highest-scoring ones have substantial room for improvement. We also present a novel suite of prompting strategies that further close the gap between LLM and human evaluators. With LLMBar, we hope to offer more insight into LLM evaluators and foster future research in developing better instruction-following models.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- Preference Ranking Optimization for Human AlignmentProceedings of the AAAI Conference on Artificial Intelligence, 2024
- Large Language Models are not Fair Evaluators2024
- Large Language Models Sensitivity to The Order of Options in Multiple-Choice Questions2024
- Instruction-following Evaluation through Verbalizer Manipulation2024
- Exploring the Use of Large Language Models for Reference-Free Text Quality Evaluation: An Empirical Study2023
- G-Eval: NLG Evaluation using Gpt-4 with Better Human Alignment2023
- Is ChatGPT a Good NLG Evaluator? A Preliminary Study2023
- Can Large Language Models Be an Alternative to Human Evaluations?2023
- Self-Instruct: Aligning Language Models with Self-Generated Instructions2023
- One Embedder, Any Task: Instruction-Finetuned Text Embeddings2023