{"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/spade-synthesizing-assertions-for-large","title":"SPADE: Synthesizing Data Quality Assertions for Large Language Model Pipelines","arxiv_id":"2401.03038","date":"2024-01-05","proceeding":null,"authors":["Shreya Shankar","Haotian Li","Parth Asawa","Madelon Hulsebos","Yiming Lin","J. D. Zamfirescu-Pereira","Harrison Chase","Will Fu-Hinthorn","Aditya G. Parameswaran","Eugene Wu"],"abstract":"Large language models (LLMs) are being increasingly deployed as part of pipelines that repeatedly process or generate data of some sort. However, a common barrier to deployment are the frequent and often unpredictable errors that plague LLMs. Acknowledging the inevitability of these errors, we propose {\\em data quality assertions} to identify when LLMs may be making mistakes. We present SPADE, a method for automatically synthesizing data quality assertions that identify bad LLM outputs. We make the observation that developers often identify data quality issues during prototyping prior to deployment, and attempt to address them by adding instructions to the LLM prompt over time. SPADE therefore analyzes histories of prompt versions over time to create candidate assertion functions and then selects a minimal set that fulfills both coverage and accuracy requirements. In testing across nine different real-world LLM pipelines, SPADE efficiently reduces the number of assertions by 14\\% and decreases false failures by 21\\% when compared to simpler baselines. SPADE has been deployed as an offering within LangSmith, LangChain's LLM pipeline hub, and has been used to generate data quality assertions for over 2000 pipelines across a spectrum of industries.","url_abs":"https://arxiv.org/abs/2401.03038v2","url_pdf":"https://arxiv.org/pdf/2401.03038v2.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":"links_only","authors_date_abstract":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license), from the Kaggle arXiv metadata snapshot of 2026-09-12"},"code_links":[{"paper_slug":"spade-synthesizing-assertions-for-large","repo_url":"https://github.com/shreyashankar/spade-experiments","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2401.03038","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2401.03038"}},"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/shreyashankar/spade-experiments","reach":{"status":"ok"}}],"summary":{"ran":15,"unverified":2},"by_repo_kind":{"official":{"samples":17,"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":17,"samples":[{"code_sha256_prefix":"7017da1e68286a10","entry":"ask_llm","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/codereviews/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/codereviews/candidate_assertions.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":"7017da1e68286a10"}},{"code_sha256_prefix":"48cde357655569c7","entry":"ask_llm","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/emails/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/emails/candidate_assertions.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":"48cde357655569c7"}},{"code_sha256_prefix":"49e1aa19c774b88a","entry":"assert_includes_code_improvements_v1","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/codereviews/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/codereviews/candidate_assertions.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":"49e1aa19c774b88a"}},{"code_sha256_prefix":"d9bcdd3a934191fc","entry":"assert_response_follows_review_format_v1","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/codereviews/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/codereviews/candidate_assertions.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":"d9bcdd3a934191fc"}},{"code_sha256_prefix":"3e08867dd69902b4","entry":"assert_structured_format_followed_1","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/emails/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/emails/candidate_assertions.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":"3e08867dd69902b4"}},{"code_sha256_prefix":"7f9b221d3ccd89d1","entry":"assert_workflow_followed_2","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"paper_experiments/emails/candidate_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/paper_experiments/emails/candidate_assertions.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":"7f9b221d3ccd89d1"}},{"code_sha256_prefix":"53f4c7efd1b7e4a0","entry":"calculate_coverage","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/optimizer.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/optimizer.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":"53f4c7efd1b7e4a0"}},{"code_sha256_prefix":"c2802c9c9076c140","entry":"calculate_ffr","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/optimizer.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/optimizer.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":"c2802c9c9076c140"}},{"code_sha256_prefix":"cfdd045ac1a9c656","entry":"does_func1_imply_func2","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/check_subsumes.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/check_subsumes.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":"cfdd045ac1a9c656"}},{"code_sha256_prefix":"451bac2fe4a2e8d3","entry":"execute_candidate_assertions","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/execute_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/execute_assertions.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":"451bac2fe4a2e8d3"}},{"code_sha256_prefix":"cea97b04cf63f5b6","entry":"generate_assertions","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/assertion_gen.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/assertion_gen.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":"cea97b04cf63f5b6"}},{"code_sha256_prefix":"1db4a844d3afd5ea","entry":"get_clean_ast","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/check_subsumes.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/check_subsumes.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":"1db4a844d3afd5ea"}},{"code_sha256_prefix":"47210ce2a56ded82","entry":"get_clean_ast_dup","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/check_subsumes.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/check_subsumes.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":"47210ce2a56ded82"}},{"code_sha256_prefix":"b7bd217d972994f3","entry":"label_responses","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/label_results.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/label_results.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":"b7bd217d972994f3"}},{"code_sha256_prefix":"c0acd8f49d961a35","entry":"prepare_for_optimizer","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/label_results.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/label_results.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":"c0acd8f49d961a35"}},{"code_sha256_prefix":"a05b1d590e92bc28","entry":"execute_pipeline_on_examples","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/execute_assertions.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/execute_assertions.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":"a05b1d590e92bc28"}},{"code_sha256_prefix":"51dbe3731b9fb8a9","entry":"show_diff","repo":"shreyashankar/spade-experiments","repo_kind":"official","path":"spade_v3/candidate_gen.py","file_url":"https://github.com/shreyashankar/spade-experiments/blob/HEAD/spade_v3/candidate_gen.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":"51dbe3731b9fb8a9"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}