AIME: AI System Optimization via Multiple LLM Evaluators

• URL: http://arxiv.org/abs/2410.03131v3
• Date: Tue, 29 Oct 2024 02:35:14 GMT
• Title: AIME: AI System Optimization via Multiple LLM Evaluators
• Authors: Bhrij Patel, Souradip Chakraborty, Wesley A. Suttle, Mengdi Wang, Amrit Singh Bedi, Dinesh Manocha,
• Abstract summary: AIME is an evaluation protocol that utilizes multiple LLMs that each independently generate an evaluation on separate criteria and then combine them via concatenation. We show AIME outperforming baseline methods in code generation tasks, with up to $62%$ higher error detection rate and up to $16%$ higher success rate than a single LLM evaluation protocol on LeetCodeHard and HumanEval datasets.
• Score: 79.03422337674664
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Text-based AI system optimization typically involves a feedback loop scheme where a single LLM generates an evaluation in natural language of the current output to improve the next iteration's output. However, in this work, we empirically demonstrate that for a practical and complex task (code generation) with multiple criteria to evaluate, utilizing only one LLM evaluator tends to let errors in generated code go undetected, thus leading to incorrect evaluations and ultimately suboptimal test case performance. Motivated by this failure case, we assume there exists an optimal evaluation policy that samples an evaluation between response and ground truth. We then theoretically prove that a linear combination of multiple evaluators can approximate this optimal policy. From this insight, we propose AI system optimization via Multiple LLM Evaluators (AIME). AIME is an evaluation protocol that utilizes multiple LLMs that each independently generate an evaluation on separate criteria and then combine them via concatenation. We provide an extensive empirical study showing AIME outperforming baseline methods in code generation tasks, with up to $62\%$ higher error detection rate and up to $16\%$ higher success rate than a single LLM evaluation protocol on LeetCodeHard and HumanEval datasets. We also show that the selection of the number of evaluators and which criteria to utilize is non-trivial as it can impact pact success rate by up to $12\%$.

Related papers

• Can LLMs Replace Human Evaluators? An Empirical Study of LLM-as-a-Judge in Software Engineering [18.766132076075365]
  Large language models (LLMs) have been deployed to tackle various software engineering (SE) tasks like code generation. Pass@k metric requires extensive unit tests and configured environments, and is not suitable for evaluating LLM-generated text. Conventional metrics like BLEU, which measure only lexical rather than semantic similarity, have also come under scrutiny.
  arXiv  Detail & Related papers  (2025-02-10T06:49:29Z)
• Re-evaluating Automatic LLM System Ranking for Alignment with Human Preference [63.03859517284341]
  An automatic evaluation framework aims to rank LLMs based on their alignment with human preferences. An automatic LLM bencher consists of four components: the input set, the evaluation model, the evaluation type and the aggregation method.
  arXiv  Detail & Related papers  (2024-12-31T17:46:51Z)
• Human-Like Code Quality Evaluation through LLM-based Recursive Semantic Comprehension [39.277408536940825]
  Code quality evaluation involves scoring generated code quality based on a reference code for a specific problem statement. Currently, there are two main forms of evaluating code quality: match-based evaluation and execution-based evaluation.
  arXiv  Detail & Related papers  (2024-11-30T01:49:25Z)
• ReIFE: Re-evaluating Instruction-Following Evaluation [105.75525154888655]
  We present a thorough meta-evaluation of instruction following, including 25 base LLMs and 15 proposed evaluation protocols. Our evaluation allows us to identify the best-performing base LLMs and evaluation protocols with a high degree of robustness.
  arXiv  Detail & Related papers  (2024-10-09T17:14:50Z)
• On Speeding Up Language Model Evaluation [48.51924035873411]
  Development of prompt-based methods with Large Language Models (LLMs) requires making numerous decisions. We propose a novel method to address this challenge. We show that it can identify the top-performing method using only 5-15% of the typically needed resources.
  arXiv  Detail & Related papers  (2024-07-08T17:48:42Z)
• How Efficient is LLM-Generated Code? A Rigorous & High-Standard Benchmark [39.13045037676502]
  Development of large language models (LLMs) has significantly pushed the frontiers of program synthesis. Most evaluation frameworks focus on the (functional) correctness of generated code; efficiency, as an important measure of code quality, has been overlooked in existing evaluations. We develop ENAMEL, a rigorous and high-standard benchmark for evaluating the capability of LLMs in generating efficient code.
  arXiv  Detail & Related papers  (2024-06-10T04:19:20Z)
• DnA-Eval: Enhancing Large Language Model Evaluation through Decomposition and Aggregation [75.81096662788254]
  Large Language Models (LLMs) are scalable and economical evaluators. The question of how reliable these evaluators are has emerged as a crucial research question. We propose Decompose and Aggregate, which breaks down the evaluation process into different stages based on pedagogical practices.
  arXiv  Detail & Related papers  (2024-05-24T08:12:30Z)
• RepEval: Effective Text Evaluation with LLM Representation [55.26340302485898]
  RepEval is a metric that leverages the projection of Large Language Models (LLMs) representations for evaluation. Our work underscores the richness of information regarding text quality embedded within LLM representations, offering insights for the development of new metrics.
  arXiv  Detail & Related papers  (2024-04-30T13:50:55Z)
• Who Validates the Validators? Aligning LLM-Assisted Evaluation of LLM Outputs with Human Preferences [11.23629471911503]
  EvalGen provides automated assistance to users in generating evaluation criteria and implementing assertions. A qualitative study finds overall support for EvalGen but underscores the subjectivity and iterative process of alignment. We identify a phenomenon we dub emphcriteria drift: users need criteria to grade outputs, but grading outputs helps users define criteria.
  arXiv  Detail & Related papers  (2024-04-18T15:45:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.