{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HPOCUIN34DMAPZGHB7Z4M2QZCZ","short_pith_number":"pith:HPOCUIN3","schema_version":"1.0","canonical_sha256":"3bdc2a21bbe0d807e4c70ff3c66a19164096c442b614b63346c53cefaf453564","source":{"kind":"arxiv","id":"2406.12824","version":1},"attestation_state":"computed","paper":{"title":"From RAGs to rich parameters: Probing how language models utilize external knowledge over parametric information for factual queries","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ehsan Aghazadeh, Hitesh Wadhwa, Rahul Seetharaman, Reshmi Ghosh, Samyadeep Basu, Shreyas Chaudhari, Somyaa Aggarwal, Soundararajan Srinivasan, Wenlong Zhao","submitted_at":"2024-06-18T17:46:08Z","abstract_excerpt":"Retrieval Augmented Generation (RAG) enriches the ability of language models to reason using external context to augment responses for a given user prompt. This approach has risen in popularity due to practical applications in various applications of language models in search, question/answering, and chat-bots. However, the exact nature of how this approach works isn't clearly understood. In this paper, we mechanistically examine the RAG pipeline to highlight that language models take shortcut and have a strong bias towards utilizing only the context information to answer the question, while r"},"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":"2406.12824","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-18T17:46:08Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"3bba1ae101f6e05ae577f359c9162e06754bf6f031b2a68b9dc65e8f6a878d2a","abstract_canon_sha256":"1f98a64cdbe0db10b7dd9c7a6056d96cea02f04c9f1d18af16cdd7f2887a3d3e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:33:51.819279Z","signature_b64":"9BSdHlgDc1lwdZkDgf8zbq64JaJdKXhDeqeVluv0i2Yzr0FvqOw3t+pwYglljv3b7owBPxufkgmQpvX+6ygaCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3bdc2a21bbe0d807e4c70ff3c66a19164096c442b614b63346c53cefaf453564","last_reissued_at":"2026-07-05T08:33:51.818795Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:33:51.818795Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From RAGs to rich parameters: Probing how language models utilize external knowledge over parametric information for factual queries","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ehsan Aghazadeh, Hitesh Wadhwa, Rahul Seetharaman, Reshmi Ghosh, Samyadeep Basu, Shreyas Chaudhari, Somyaa Aggarwal, Soundararajan Srinivasan, Wenlong Zhao","submitted_at":"2024-06-18T17:46:08Z","abstract_excerpt":"Retrieval Augmented Generation (RAG) enriches the ability of language models to reason using external context to augment responses for a given user prompt. This approach has risen in popularity due to practical applications in various applications of language models in search, question/answering, and chat-bots. However, the exact nature of how this approach works isn't clearly understood. In this paper, we mechanistically examine the RAG pipeline to highlight that language models take shortcut and have a strong bias towards utilizing only the context information to answer the question, while r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12824","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/2406.12824/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":"2406.12824","created_at":"2026-07-05T08:33:51.818851+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.12824v1","created_at":"2026-07-05T08:33:51.818851+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12824","created_at":"2026-07-05T08:33:51.818851+00:00"},{"alias_kind":"pith_short_12","alias_value":"HPOCUIN34DMA","created_at":"2026-07-05T08:33:51.818851+00:00"},{"alias_kind":"pith_short_16","alias_value":"HPOCUIN34DMAPZGH","created_at":"2026-07-05T08:33:51.818851+00:00"},{"alias_kind":"pith_short_8","alias_value":"HPOCUIN3","created_at":"2026-07-05T08:33:51.818851+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.05594","citing_title":"The Cost of Context: Mitigating Textual Bias in Multimodal Retrieval-Augmented Generation","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ","json":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ.json","graph_json":"https://pith.science/api/pith-number/HPOCUIN34DMAPZGHB7Z4M2QZCZ/graph.json","events_json":"https://pith.science/api/pith-number/HPOCUIN34DMAPZGHB7Z4M2QZCZ/events.json","paper":"https://pith.science/paper/HPOCUIN3"},"agent_actions":{"view_html":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ","download_json":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ.json","view_paper":"https://pith.science/paper/HPOCUIN3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.12824&json=true","fetch_graph":"https://pith.science/api/pith-number/HPOCUIN34DMAPZGHB7Z4M2QZCZ/graph.json","fetch_events":"https://pith.science/api/pith-number/HPOCUIN34DMAPZGHB7Z4M2QZCZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ/action/storage_attestation","attest_author":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ/action/author_attestation","sign_citation":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ/action/citation_signature","submit_replication":"https://pith.science/pith/HPOCUIN34DMAPZGHB7Z4M2QZCZ/action/replication_record"}},"created_at":"2026-07-05T08:33:51.818851+00:00","updated_at":"2026-07-05T08:33:51.818851+00:00"}