PARIKSHA: A Large-Scale Investigation of Human-LLM Evaluator Agreement on Multilingual and Multi-Cultural Data

• URL: http://arxiv.org/abs/2406.15053v2
• Date: Fri, 18 Oct 2024 08:51:55 GMT
• Title: PARIKSHA: A Large-Scale Investigation of Human-LLM Evaluator Agreement on Multilingual and Multi-Cultural Data
• Authors: Ishaan Watts, Varun Gumma, Aditya Yadavalli, Vivek Seshadri, Manohar Swaminathan, Sunayana Sitaram,
• Abstract summary: We evaluate 30 models across 10 Indic languages by conducting 90K human evaluations and 30K LLM-based evaluations. We find that models such as GPT-4o and Llama-3 70B consistently perform best for most Indic languages.
• Score: 12.852628521840542
• License:
• Abstract: Evaluation of multilingual Large Language Models (LLMs) is challenging due to a variety of factors -- the lack of benchmarks with sufficient linguistic diversity, contamination of popular benchmarks into LLM pre-training data and the lack of local, cultural nuances in translated benchmarks. In this work, we study human and LLM-based evaluation in a multilingual, multi-cultural setting. We evaluate 30 models across 10 Indic languages by conducting 90K human evaluations and 30K LLM-based evaluations and find that models such as GPT-4o and Llama-3 70B consistently perform best for most Indic languages. We build leaderboards for two evaluation settings - pairwise comparison and direct assessment and analyze the agreement between humans and LLMs. We find that humans and LLMs agree fairly well in the pairwise setting but the agreement drops for direct assessment evaluation especially for languages such as Bengali and Odia. We also check for various biases in human and LLM-based evaluation and find evidence of self-bias in the GPT-based evaluator. Our work presents a significant step towards scaling up multilingual evaluation of LLMs.

Related papers

• MM-Eval: A Multilingual Meta-Evaluation Benchmark for LLM-as-a-Judge and Reward Models [3.961168847961322]
  Large language models (LLMs) are commonly used as evaluators in tasks, where they act as proxies for human preferences or judgments. Existing benchmarks primarily focus on English, offering limited insight into LLMs' effectiveness as evaluators in non-English contexts. We introduce MM-Eval, a multilingual meta-evaluation benchmark that covers 18 languages across six categories.
  arXiv  Detail & Related papers  (2024-10-23T06:04:55Z)
• Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs [36.30321941154582]
  Hercule is a cross-lingual evaluation model that learns to assign scores to responses based on easily available reference answers in English. This study is the first comprehensive examination of cross-lingual evaluation using LLMs, presenting a scalable and effective approach for multilingual assessment.
  arXiv  Detail & Related papers  (2024-10-17T09:45:32Z)
• On the Evaluation Practices in Multilingual NLP: Can Machine Translation Offer an Alternative to Human Translations? [19.346078451375693]
  We present an analysis of existing evaluation frameworks in NLP. We propose several directions for more robust and reliable evaluation practices. We show that simpler baselines can achieve relatively strong performance without having benefited from large-scale multilingual pretraining.
  arXiv  Detail & Related papers  (2024-06-20T12:46:12Z)
• Quantifying Multilingual Performance of Large Language Models Across Languages [48.40607157158246]
  Large Language Models (LLMs) perform better on high-resource languages like English, German, and French, while their capabilities in low-resource languages remain inadequate. We propose the Language Ranker, an intrinsic metric designed to benchmark and rank languages based on LLM performance using internal representations. Our analysis reveals that high-resource languages exhibit higher similarity scores with English, demonstrating superior performance, while low-resource languages show lower similarity scores.
  arXiv  Detail & Related papers  (2024-04-17T16:53:16Z)
• METAL: Towards Multilingual Meta-Evaluation [12.852595634767901]
  This study proposes a framework for an end-to-end assessment of Large Language Models (LLMs) as evaluators in multilingual scenarios. We create a dataset covering 10 languages containing native speaker judgments for the task of summarization. We compare the performance of LLM-based evaluators created using GPT-3.5-Turbo, GPT-4, and PaLM2.
  arXiv  Detail & Related papers  (2024-04-02T06:14:54Z)
• OMGEval: An Open Multilingual Generative Evaluation Benchmark for Large Language Models [59.54423478596468]
  We introduce OMGEval, the first Open-source Multilingual Generative test set that can assess the capability of LLMs in different languages. For each language, OMGEval provides 804 open-ended questions, covering a wide range of important capabilities of LLMs. Specifically, the current version of OMGEval includes 5 languages (i.e., Zh, Ru, Fr, Es, Ar)
  arXiv  Detail & Related papers  (2024-02-21T04:42:41Z)
• Are Large Language Model-based Evaluators the Solution to Scaling Up Multilingual Evaluation? [20.476500441734427]
  Large Language Models (LLMs) excel in various Natural Language Processing (NLP) tasks. Their evaluation, particularly in languages beyond the top $20$, remains inadequate due to existing benchmarks and metrics limitations.
  arXiv  Detail & Related papers  (2023-09-14T06:41:58Z)
• L-Eval: Instituting Standardized Evaluation for Long Context Language Models [91.05820785008527]
  We propose L-Eval to institute a more standardized evaluation for long context language models (LCLMs) We build a new evaluation suite containing 20 sub-tasks, 508 long documents, and over 2,000 human-labeled query-response pairs. Results show that popular n-gram matching metrics generally can not correlate well with human judgment.
  arXiv  Detail & Related papers  (2023-07-20T17:59:41Z)
• Evaluating the Performance of Large Language Models on GAOKAO Benchmark [53.663757126289795]
  This paper introduces GAOKAO-Bench, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples. With human evaluation, we obtain the converted total score of LLMs, including GPT-4, ChatGPT and ERNIE-Bot. We also use LLMs to grade the subjective questions, and find that model scores achieve a moderate level of consistency with human scores.
  arXiv  Detail & Related papers  (2023-05-21T14:39:28Z)
• Can Large Language Models Be an Alternative to Human Evaluations? [80.81532239566992]
  Large language models (LLMs) have demonstrated exceptional performance on unseen tasks when only the task instructions are provided. We show that the result of LLM evaluation is consistent with the results obtained by expert human evaluation.
  arXiv  Detail & Related papers  (2023-05-03T07:28:50Z)
• Benchmarking Large Language Models for News Summarization [79.37850439866938]
  Large language models (LLMs) have shown promise for automatic summarization but the reasons behind their successes are poorly understood. We find instruction tuning, and not model size, is the key to the LLM's zero-shot summarization capability.
  arXiv  Detail & Related papers  (2023-01-31T18:46:19Z)

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.