RECKON: Large-scale Reference-based Efficient Knowledge Evaluation for Large Language Model

Lin Zhang, Zhouhong Gu, Xiaoran Shi, Hongwei Feng, Yanghua Xiao
4/1/2025
cs.CL

Official code: MikeGu721/reckon (0 stars, last commit 2025-04-02).

Abstract

As large language models (LLMs) advance, efficient knowledge evaluation becomes crucial to verifying their capabilities. Traditional methods, relying on benchmarks, face limitations such as high resource costs and information loss. We propose the Large-scale Reference-based Efficient Knowledge Evaluation for Large Language Model (RECKON), which directly uses reference data to evaluate models. RECKON organizes unstructured data into manageable units and generates targeted questions for each cluster, improving evaluation accuracy and efficiency. Experimental results show that RECKON reduces resource consumption by 56.5% compared to traditional methods while achieving over 97% accuracy across various domains, including world knowledge, code, legal, and biomedical datasets. Code is available at https://github.com/MikeGu721/reckon

View on arXivView PDF

Code Implementations(1)

MikeGu721/reckonOfficial100%
00PythonApr 1, 20251 years ago

Similar Papers

Estimating Knowledge in Large Language Models Without Generating a Single Token

84% match2024

SciKnowEval: Evaluating Multi-level Scientific Knowledge of Large Language Models

84% match1 repo282024

KIEval: A Knowledge-grounded Interactive Evaluation Framework for Large Language Models

83% match2 repos72.8k2024

KoLA: Carefully Benchmarking World Knowledge of Large Language Models

83% match1 repo572026

Benchmarking Knowledge Boundary for Large Language Models: A Different Perspective on Model Evaluation

83% match1 repo102024

Survey on Knowledge Distillation for Large Language Models: Methods, Evaluation, and Application

83% match2024

Cite this paper

@article{zhang2025reckon,
  title  = {RECKON: Large-scale Reference-based Efficient Knowledge Evaluation for Large Language Model},
  author = {Lin Zhang and Zhouhong Gu and Xiaoran Shi and Hongwei Feng and Yanghua Xiao},
  year   = {2025},
  eprint = {2504.00756},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2504.00756v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2504.00756.svg)](https://codeofpaper.com/paper/2504.00756)

Discussion