{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/an-empirical-study-of-llm-as-a-judge-for-llm","title":"An Empirical Study of LLM-as-a-Judge for LLM Evaluation: Fine-tuned Judge Model is not a General Substitute for GPT-4","arxiv_id":"2403.02839","date":"2024-03-05","proceeding":null,"authors":["Hui Huang","Yingqi Qu","Xingyuan Bu","Hongli Zhou","Jing Liu","Muyun Yang","Bing Xu","Tiejun Zhao"],"abstract":"Recently, there has been a growing trend of utilizing Large Language Model (LLM) to evaluate the quality of other LLMs. Many studies have employed proprietary close-sourced models, especially GPT-4, as the evaluator. Alternatively, other works have fine-tuned judge models based on open-source LLMs as the evaluator. While the fine-tuned judge models are claimed to achieve comparable evaluation capability with GPT-4, in this work, we conduct an empirical study of judge models. Our findings indicate that although the fine-tuned judge models achieve high performance on in-domain test sets, even surpassing GPT-4, they underperform GPT-4 across several dimensions, including generalizability, fairness, aspect-specific evaluation, and scalability. We also reveal that the fine-tuned judge model inherently operates as a task-specific classifier, consequently imposing the limitations. Finally, we introduce a integrated method, leveraging GPT-4 to compensate for the limitations and improve the fine-tuned judges. Experiment results show our method achieves accuracy on par with GPT-4 with only 50% of the API expense.","url_abs":"https://arxiv.org/abs/2403.02839v3","url_pdf":"https://arxiv.org/pdf/2403.02839v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"an-empirical-study-of-llm-as-a-judge-for-llm","repo_url":"https://github.com/huihuichyan/unlimitedjudge","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"fairness","task_name":"Fairness"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"large-language-model","task_name":"Large Language Model"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"gpt-4","method_name":"GPT-4"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2403.02839","atlas_url":"https://app.syntology.ai/?focus=2403.02839","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2403.02839"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/huihuichyan/unlimitedjudge","reach":{"status":"ok"}}],"summary":{"ran":15,"unverified":3},"by_repo_kind":{"official":{"samples":18,"ran":15,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":18,"samples":[{"code_sha256_prefix":"08ef47c526ea281d","entry":"build_dataset","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_dataset.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_dataset.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"08ef47c526ea281d"}},{"code_sha256_prefix":"ba3dc13d4a10674d","entry":"build_demo_instruction","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_icl.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_icl.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ba3dc13d4a10674d"}},{"code_sha256_prefix":"acacb923208da24f","entry":"calculate_metrics","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_dataset.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_dataset.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"acacb923208da24f"}},{"code_sha256_prefix":"5bb3a6f55ea7c8a7","entry":"create_prompt","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_prompt_judge.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_prompt_judge.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"5bb3a6f55ea7c8a7"}},{"code_sha256_prefix":"b3735f4c46bf4dc2","entry":"create_prompt_cot","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_prompt_judge.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_prompt_judge.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b3735f4c46bf4dc2"}},{"code_sha256_prefix":"610765b37afa23c4","entry":"create_prompt_gpt","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_prompt_gpt.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_prompt_gpt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"610765b37afa23c4"}},{"code_sha256_prefix":"6df9dfc27620344a","entry":"format_instruction","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/finetune.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/finetune.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6df9dfc27620344a"}},{"code_sha256_prefix":"8588c2345ea473e0","entry":"get_average_scores","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/cascaded_eval.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/cascaded_eval.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8588c2345ea473e0"}},{"code_sha256_prefix":"34391b337ca8d626","entry":"get_batch_evaluation","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/cal_reliability.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/cal_reliability.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"34391b337ca8d626"}},{"code_sha256_prefix":"c7c6421e96f43d9d","entry":"load_results","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/cascaded_eval.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/cascaded_eval.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c7c6421e96f43d9d"}},{"code_sha256_prefix":"972a8f7586d01eca","entry":"normalize_scores","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/evaluate_reliability.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/evaluate_reliability.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"972a8f7586d01eca"}},{"code_sha256_prefix":"da11ac6a2f3ad115","entry":"parse_predictions","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_prompt_judge.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_prompt_judge.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"da11ac6a2f3ad115"}},{"code_sha256_prefix":"9d69b5ab24ef072e","entry":"parse_score_gpt","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_prompt_gpt.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_prompt_gpt.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9d69b5ab24ef072e"}},{"code_sha256_prefix":"b0ea70c48e350289","entry":"select_top_half_indices","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/cascaded_eval.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/cascaded_eval.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b0ea70c48e350289"}},{"code_sha256_prefix":"256c39e7e8b2a35b","entry":"swap_first_two_integers","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/finetune.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/finetune.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"256c39e7e8b2a35b"}},{"code_sha256_prefix":"09f28e6be3e79147","entry":"batched_classification","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/evaluate_finetuned.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/evaluate_finetuned.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"09f28e6be3e79147"}},{"code_sha256_prefix":"06e352b4b97e89ac","entry":"build_icl","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_icl.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_icl.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"06e352b4b97e89ac"}},{"code_sha256_prefix":"e152cae199d3f95b","entry":"build_trainset","repo":"huihuichyan/unlimitedjudge","repo_kind":"official","path":"src/build_icl.py","file_url":"https://github.com/huihuichyan/unlimitedjudge/blob/HEAD/src/build_icl.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e152cae199d3f95b"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}