{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FO622MPUJCX7WRK5SY3SVHWFFJ","short_pith_number":"pith:FO622MPU","schema_version":"1.0","canonical_sha256":"2bbdad31f448affb455d96372a9ec52a4713f902c1a3090ab505e119c54667f9","source":{"kind":"arxiv","id":"2401.01701","version":3},"attestation_state":"computed","paper":{"title":"De-Hallucinator: Mitigating LLM Hallucinations in Code Generation Tasks via Iterative Grounding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Aryaz Eghbali, Michael Pradel","submitted_at":"2024-01-03T12:09:43Z","abstract_excerpt":"Large language models (LLMs) trained on datasets of publicly available source code have established a new state of the art in code generation tasks. However, these models are mostly unaware of the code that exists within a specific project, preventing the models from making good use of existing APIs. Instead, LLMs often invent, or \"hallucinate\", non-existent APIs or produce variants of already existing code. This paper presents De-Hallucinator, a technique that grounds the predictions of an LLM through a novel combination of retrieving suitable API references and iteratively querying the model"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2401.01701","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-01-03T12:09:43Z","cross_cats_sorted":[],"title_canon_sha256":"37a71ebc77839b62de2968bb6d31e5ed8b35fe1d0cb825459aad6fda727b561d","abstract_canon_sha256":"dc3fd74dc2f028b2ac58464579f88c9a910d32bc9740fb19302c4c778272f3e8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:34:09.561845Z","signature_b64":"JPBF8yDv8vrHFKvFi06ZRXQWj5PmarlLMwdk4l1XusLb1LDk28+Vg0iKQNdjsBWQh70SkE0hoJGwwHr1M0OUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2bbdad31f448affb455d96372a9ec52a4713f902c1a3090ab505e119c54667f9","last_reissued_at":"2026-07-05T08:34:09.561357Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:34:09.561357Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"De-Hallucinator: Mitigating LLM Hallucinations in Code Generation Tasks via Iterative Grounding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SE","authors_text":"Aryaz Eghbali, Michael Pradel","submitted_at":"2024-01-03T12:09:43Z","abstract_excerpt":"Large language models (LLMs) trained on datasets of publicly available source code have established a new state of the art in code generation tasks. However, these models are mostly unaware of the code that exists within a specific project, preventing the models from making good use of existing APIs. Instead, LLMs often invent, or \"hallucinate\", non-existent APIs or produce variants of already existing code. This paper presents De-Hallucinator, a technique that grounds the predictions of an LLM through a novel combination of retrieving suitable API references and iteratively querying the model"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.01701","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2401.01701/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2401.01701","created_at":"2026-07-05T08:34:09.561416+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.01701v3","created_at":"2026-07-05T08:34:09.561416+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.01701","created_at":"2026-07-05T08:34:09.561416+00:00"},{"alias_kind":"pith_short_12","alias_value":"FO622MPUJCX7","created_at":"2026-07-05T08:34:09.561416+00:00"},{"alias_kind":"pith_short_16","alias_value":"FO622MPUJCX7WRK5","created_at":"2026-07-05T08:34:09.561416+00:00"},{"alias_kind":"pith_short_8","alias_value":"FO622MPU","created_at":"2026-07-05T08:34:09.561416+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.05810","citing_title":"SCOPE: Leveraging Subgoal Critiques for Code Generation","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"2607.05936","citing_title":"Mitigating Errors in LLM-Generated Web API Invocations via Retrieval-Augmented Generation and Constrained Decoding","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2606.09800","citing_title":"FASE: Fast Adaptive Semantic Entropy for Code Quality","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08135","citing_title":"TICoder: A Repository-Level Code Generation Framework with Test-Driven Planning and Implementation-Aware Reuse","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29372","citing_title":"On the Road to Personalized Code Intelligence: Portraiting and Assisting Developers Based on Their In-IDE Behaviors","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2505.04080","citing_title":"MojoFrame: Dataframe Library in Mojo Language","ref_index":128,"is_internal_anchor":false},{"citing_arxiv_id":"2408.15815","citing_title":"MR-Adopt: Automatic Deduction of Input Transformation Function for Metamorphic Testing","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22202","citing_title":"Library Hallucinations in LLM-Generated Code: A Risk Analysis Grounded in Developer Queries","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17029","citing_title":"Task Abstention for Large Language Models in Code Generation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15503","citing_title":"uGen: An Agentic Framework for Generating Microarchitectural Attack PoCs","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2403.17134","citing_title":"RepairAgent: An Autonomous, LLM-Based Agent for Program Repair","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2603.06276","citing_title":"Story Point Estimation Using Large Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2402.19473","citing_title":"Retrieval-Augmented Generation for AI-Generated Content: A Survey","ref_index":241,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26672","citing_title":"What Makes Software Bugs Escape Testing? Evidence from a Large-Scale Empirical Study","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10126","citing_title":"MR-Coupler: Automated Metamorphic Test Generation via Functional Coupling Analysis","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ","json":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ.json","graph_json":"https://pith.science/api/pith-number/FO622MPUJCX7WRK5SY3SVHWFFJ/graph.json","events_json":"https://pith.science/api/pith-number/FO622MPUJCX7WRK5SY3SVHWFFJ/events.json","paper":"https://pith.science/paper/FO622MPU"},"agent_actions":{"view_html":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ","download_json":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ.json","view_paper":"https://pith.science/paper/FO622MPU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.01701&json=true","fetch_graph":"https://pith.science/api/pith-number/FO622MPUJCX7WRK5SY3SVHWFFJ/graph.json","fetch_events":"https://pith.science/api/pith-number/FO622MPUJCX7WRK5SY3SVHWFFJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ/action/storage_attestation","attest_author":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ/action/author_attestation","sign_citation":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ/action/citation_signature","submit_replication":"https://pith.science/pith/FO622MPUJCX7WRK5SY3SVHWFFJ/action/replication_record"}},"created_at":"2026-07-05T08:34:09.561416+00:00","updated_at":"2026-07-05T08:34:09.561416+00:00"}