{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NI2OKBAFJINEWEHD6ECGFL6CGF","short_pith_number":"pith:NI2OKBAF","schema_version":"1.0","canonical_sha256":"6a34e504054a1a4b10e3f10462afc2317993c49dcdb41b7513311aebcb254649","source":{"kind":"arxiv","id":"2506.17277","version":1},"attestation_state":"computed","paper":{"title":"Chunk Twice, Embed Once: A Systematic Study of Segmentation and Representation Trade-offs in Chemistry-Aware Retrieval-Augmented Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","physics.chem-ph"],"primary_cat":"cs.IR","authors_text":"Mahmoud Amiri, Thomas Bocklitz","submitted_at":"2025-06-13T07:44:53Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) systems are increasingly vital for navigating the ever-expanding body of scientific literature, particularly in high-stakes domains such as chemistry. Despite the promise of RAG, foundational design choices -- such as how documents are segmented and represented -- remain underexplored in domain-specific contexts. This study presents the first large-scale, systematic evaluation of chunking strategies and embedding models tailored to chemistry-focused RAG systems. We investigate 25 chunking configurations across five method families and evaluate 48 embedding "},"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":"2506.17277","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2025-06-13T07:44:53Z","cross_cats_sorted":["cs.AI","physics.chem-ph"],"title_canon_sha256":"30fb06b4614138158332377eaa48fce370ead0698a9b26975e6150328ec7a597","abstract_canon_sha256":"5d0fe186335200335d637094ff772caae120372d68bcaf5c79ddbb06b3dfea0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:14.686794Z","signature_b64":"0DsvK+RkhGVoJDC+ysXy792HsRkqIgqlo2s0k5zIgu5vUHeBvR7N4fj4hjIkEWBZtabcpdgn3Lz3ryf1EJg4Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6a34e504054a1a4b10e3f10462afc2317993c49dcdb41b7513311aebcb254649","last_reissued_at":"2026-07-05T11:25:14.686304Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:14.686304Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Chunk Twice, Embed Once: A Systematic Study of Segmentation and Representation Trade-offs in Chemistry-Aware Retrieval-Augmented Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","physics.chem-ph"],"primary_cat":"cs.IR","authors_text":"Mahmoud Amiri, Thomas Bocklitz","submitted_at":"2025-06-13T07:44:53Z","abstract_excerpt":"Retrieval-Augmented Generation (RAG) systems are increasingly vital for navigating the ever-expanding body of scientific literature, particularly in high-stakes domains such as chemistry. Despite the promise of RAG, foundational design choices -- such as how documents are segmented and represented -- remain underexplored in domain-specific contexts. This study presents the first large-scale, systematic evaluation of chunking strategies and embedding models tailored to chemistry-focused RAG systems. We investigate 25 chunking configurations across five method families and evaluate 48 embedding "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.17277","kind":"arxiv","version":1},"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/2506.17277/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":"2506.17277","created_at":"2026-07-05T11:25:14.686370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.17277v1","created_at":"2026-07-05T11:25:14.686370+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.17277","created_at":"2026-07-05T11:25:14.686370+00:00"},{"alias_kind":"pith_short_12","alias_value":"NI2OKBAFJINE","created_at":"2026-07-05T11:25:14.686370+00:00"},{"alias_kind":"pith_short_16","alias_value":"NI2OKBAFJINEWEHD","created_at":"2026-07-05T11:25:14.686370+00:00"},{"alias_kind":"pith_short_8","alias_value":"NI2OKBAF","created_at":"2026-07-05T11:25:14.686370+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.12498","citing_title":"Lit2Vec: A Reproducible Workflow for Building a Legally Screened Chemistry Corpus from S2ORC for Downstream Retrieval and Text Mining","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF","json":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF.json","graph_json":"https://pith.science/api/pith-number/NI2OKBAFJINEWEHD6ECGFL6CGF/graph.json","events_json":"https://pith.science/api/pith-number/NI2OKBAFJINEWEHD6ECGFL6CGF/events.json","paper":"https://pith.science/paper/NI2OKBAF"},"agent_actions":{"view_html":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF","download_json":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF.json","view_paper":"https://pith.science/paper/NI2OKBAF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.17277&json=true","fetch_graph":"https://pith.science/api/pith-number/NI2OKBAFJINEWEHD6ECGFL6CGF/graph.json","fetch_events":"https://pith.science/api/pith-number/NI2OKBAFJINEWEHD6ECGFL6CGF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF/action/storage_attestation","attest_author":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF/action/author_attestation","sign_citation":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF/action/citation_signature","submit_replication":"https://pith.science/pith/NI2OKBAFJINEWEHD6ECGFL6CGF/action/replication_record"}},"created_at":"2026-07-05T11:25:14.686370+00:00","updated_at":"2026-07-05T11:25:14.686370+00:00"}