{"paper":{"title":"[Experiment, Analysis, and Benchmark] How well do LLMs comprehend and design relational database schemata?","arxiv_id":"2611.17044","primary_cat":"cs.DB","submitted_at":"2026-11-12T16:28:21+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":0.83,"id":"s0","own":false,"role":"background","text":"Large Language Models (LLMs) are increasingly used for database tasks, yet most work, such as Text2SQL, assumes that the relational schema already exists.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.06,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.2,"role_probabilities":{"gap":0.17,"method":0.0,"result":0.0,"background":0.83,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.75,"id":"s1","own":false,"role":"contribution","text":"We ask whether LLMs can design a schema from natural language requirements (Text2Schema), verbalize an existing schema (Schema2Text), and preserve its structure through a Schema-to-Text-to-Schema round trip.","hedge":{"p":0.55,"label":"unclear"},"own_p":0.29,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.04,"role_probabilities":{"gap":0.23,"method":0.0,"result":0.0,"background":0.02,"limitation":0.0,"future_work":0.0,"contribution":0.75,"interpretation":0.0}},{"i":2,"p":0.57,"id":"s2","own":true,"role":"method","text":"We present SchemaBridge, a benchmark of 902 schemata from three collections, stratified by size and constraint complexity, that scores fidelity separately for tables, attributes, domain (data types and nullability), entity, and referential integrity.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.95,"main_claim_p":0.04,"quantitative":true,"quantitative_p":0.97,"role_probabilities":{"gap":0.0,"method":0.57,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.43,"interpretation":0.0}},{"i":3,"p":1.0,"id":"s3","own":true,"role":"result","text":"Evaluating six open-weight and proprietary LLMs, we find that: (i) human-written requirements under-specify schemata, with no model exceeding 0.27 referential integrity F1; (ii) LLM-generated descriptions carry far more structure when explicitly asked to be complete, raising GPT-5.4-mini's referential F1 on SchemaBench from 0.228 to 0.909, even as BLEU, ROUGE-L, and BERTScore decrease; (iii) controlled experiments show that both constructor and describer choices influence round-trip fidelity under fixed settings; and (iv) masked schema experiments show that structural cues support recovery of keys and relationships, while joint data-type and nullability information is poorly retained through verbalization.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.98,"main_claim_p":0.83,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":4,"p":0.85,"id":"s4","own":false,"role":"contribution","text":"Code and data are available at https://github.com/iamsriom/SchemaBridge","hedge":{"p":0.83,"label":"asserted"},"own_p":0.3,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.03,"result":0.0,"background":0.01,"limitation":0.0,"future_work":0.11,"contribution":0.85,"interpretation":0.0}}]},"compiled":{"main_claim":{"p":1.0,"text":"Evaluating six open-weight and proprietary LLMs, we find that: (i) human-written requirements under-specify schemata, with no model exceeding 0.27 referential integrity F1; (ii) LLM-generated descriptions carry far more structure when explicitly asked to be complete, raising GPT-5.4-mini's referential F1 on SchemaBench from 0.228 to 0.909, even as BLEU, ROUGE-L, and BERTScore decrease; (iii) controlled experiments show that both constructor and describer choices influence round-trip fidelity under fixed settings; and (iv) masked schema experiments show that structural cues support recovery of keys and relationships, while joint data-type and nullability information is poorly retained through verbalization.","location":"abstract s3","hedge_strength":{"p":1.0,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":4592,"cost_usd_reported":0.0},"compiled_at":"2026-11-13","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.08,"states_limitation":0.85,"human_participants":0.11,"controlled_comparison":0.54,"randomized_assignment":0.02,"states_quantitative_result":0.99,"compares_with_prior_results":0.57,"mentions_code_or_data_availability":0.99},"evidential_basis":{"p":1.0,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.0,"observation_or_experiment":1.0,"simulation_or_computation":0.0}},"contribution_type":{"p":0.99,"value":"dataset or benchmark","probabilities":{"review":0.0,"method_or_tool":0.0,"empirical_study":0.01,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.99,"position_or_commentary":0.0}},"contribution_statement":{"p":0.96,"text":"We present SchemaBridge, a benchmark of 902 schemata from three collections, stratified by size and constraint complexity, that scores fidelity separately for tables, attributes, domain (data types and nullability), entity, and referential integrity.","location":"abstract s2"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":5,"sentences_in_abstract":5},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}