{"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/factscore-fine-grained-atomic-evaluation-of","title":"FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation","arxiv_id":"2305.14251","date":"2023-05-23","proceeding":null,"authors":["Sewon Min","Kalpesh Krishna","Xinxi Lyu","Mike Lewis","Wen-tau Yih","Pang Wei Koh","Mohit Iyyer","Luke Zettlemoyer","Hannaneh Hajishirzi"],"abstract":"Evaluating the factuality of long-form text generated by large language models (LMs) is non-trivial because (1) generations often contain a mixture of supported and unsupported pieces of information, making binary judgments of quality inadequate, and (2) human evaluation is time-consuming and costly. In this paper, we introduce FACTSCORE, a new evaluation that breaks a generation into a series of atomic facts and computes the percentage of atomic facts supported by a reliable knowledge source. We conduct an extensive human evaluation to obtain FACTSCOREs of people biographies generated by several state-of-the-art commercial LMs -- InstructGPT, ChatGPT, and the retrieval-augmented PerplexityAI -- and report new analysis demonstrating the need for such a fine-grained score (e.g., ChatGPT only achieves 58%). Since human evaluation is costly, we also introduce an automated model that estimates FACTSCORE using retrieval and a strong language model, with less than a 2% error rate. Finally, we use this automated metric to evaluate 6,500 generations from a new set of 13 recent LMs that would have cost $26K if evaluated by humans, with various findings: GPT-4 and ChatGPT are more factual than public models, and Vicuna and Alpaca are some of the best public models. FACTSCORE is available for public use via `pip install factscore`.","url_abs":"https://arxiv.org/abs/2305.14251v2","url_pdf":"https://arxiv.org/pdf/2305.14251v2.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":"factscore-fine-grained-atomic-evaluation-of","repo_url":"https://github.com/shmsw25/factscore","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"factscore-fine-grained-atomic-evaluation-of","repo_url":"https://github.com/d223302/merging-facts-crafting-fallacies","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"factscore-fine-grained-atomic-evaluation-of","repo_url":"https://github.com/freshllms/freshqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"factscore-fine-grained-atomic-evaluation-of","repo_url":"https://github.com/tatsu-lab/linguistic_calibration","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"form","task_name":"Form"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"text-generation","task_name":"Text Generation"}],"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":{"atlas_url":"https://app.syntology.ai/?focus=2305.14251","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.14251"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/shmsw25/factscore","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/freshllms/freshqa","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/d223302/merging-facts-crafting-fallacies","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tatsu-lab/linguistic_calibration","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/shmsw25/FActScore","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":9,"ran_draft_wrong":1,"unverified":5},"by_repo_kind":{"official":{"samples":12,"ran":9,"repositories":1},"listed":{"samples":3,"ran":1,"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":0,"samples":[{"code_sha256_prefix":"7ad7a356ddeef1cb","entry":"Retrieval","repo":"tatsu-lab/linguistic_calibration","repo_kind":"listed","path":"src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","file_url":"https://github.com/tatsu-lab/linguistic_calibration/blob/HEAD/src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"7ad7a356ddeef1cb"}},{"code_sha256_prefix":"510528f4db1006bc","entry":"best_demos","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/atomic_facts.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/atomic_facts.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"510528f4db1006bc"}},{"code_sha256_prefix":"6c0eab0d68ea3036","entry":"call_ChatGPT","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/openai_lm.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/openai_lm.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6c0eab0d68ea3036"}},{"code_sha256_prefix":"e5bb87d92ea99a8b","entry":"call_GPT3","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/openai_lm.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/openai_lm.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e5bb87d92ea99a8b"}},{"code_sha256_prefix":"f06c53e363e1ebaa","entry":"get_memory_footprint","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/utils.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/utils.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f06c53e363e1ebaa"}},{"code_sha256_prefix":"8eb846a8e76a79b8","entry":"is_invalid_paragraph_ppl","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/abstain_detection.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/abstain_detection.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8eb846a8e76a79b8"}},{"code_sha256_prefix":"672bd38d830be48d","entry":"is_invalid_ppl","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/abstain_detection.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/abstain_detection.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"672bd38d830be48d"}},{"code_sha256_prefix":"6a96435eba311b08","entry":"normalize_answer","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/atomic_facts.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/atomic_facts.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6a96435eba311b08"}},{"code_sha256_prefix":"3635046105ae0fce","entry":"remove_citation","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/abstain_detection.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/abstain_detection.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3635046105ae0fce"}},{"code_sha256_prefix":"13867f872a481bd3","entry":"text_to_sentences","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/atomic_facts.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/atomic_facts.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"13867f872a481bd3"}},{"code_sha256_prefix":"c4456232ee580c04","entry":"DocDB","repo":"tatsu-lab/linguistic_calibration","repo_kind":"listed","path":"src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","file_url":"https://github.com/tatsu-lab/linguistic_calibration/blob/HEAD/src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"c4456232ee580c04"}},{"code_sha256_prefix":"d14a268e400630cf","entry":"FactScorer","repo":"tatsu-lab/linguistic_calibration","repo_kind":"listed","path":"src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","file_url":"https://github.com/tatsu-lab/linguistic_calibration/blob/HEAD/src/linguistic_calibration/auto_annotations/factscore_retrieval_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d14a268e400630cf"}},{"code_sha256_prefix":"7e4c366c4dbf9568","entry":"convert_model_to_int8_on_gpu","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/utils.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/utils.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7e4c366c4dbf9568"}},{"code_sha256_prefix":"02c83d89394be1c7","entry":"recover_instruct_llama","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/download_data.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/download_data.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"02c83d89394be1c7"}},{"code_sha256_prefix":"925c59b7e44138ab","entry":"softmax","repo":"shmsw25/factscore","repo_kind":"official","path":"factscore/npm.py","file_url":"https://github.com/shmsw25/factscore/blob/HEAD/factscore/npm.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"925c59b7e44138ab"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}