{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CG22D2SSU6Z62BEZGKTZFG3M6X","short_pith_number":"pith:CG22D2SS","schema_version":"1.0","canonical_sha256":"11b5a1ea52a7b3ed049932a7929b6cf5e71b1d6e43c56eed47329a1ed24a9d78","source":{"kind":"arxiv","id":"2601.16651","version":3},"attestation_state":"computed","paper":{"title":"Select or Project? Evaluating Lower-dimensional Vectors for LLM Training Data Explanations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin Roth, Loris Schoenegger, Lukas Hinterleitner","submitted_at":"2026-01-23T11:15:20Z","abstract_excerpt":"Gradient-based methods for instance-based explanation for large language models (LLMs) are hindered by the immense dimensionality of model gradients. In practice, influence estimation is restricted to a subset of model parameters to make computation tractable, but this subset is often chosen ad hoc and rarely justified by systematic evaluation. This paper investigates if it is better to create low-dimensional representations by selecting a small, architecturally informed subset of model components or by projecting the full gradients into a lower-dimensional space. Using a novel benchmark, we s"},"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":"2601.16651","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-23T11:15:20Z","cross_cats_sorted":[],"title_canon_sha256":"edbdd8812257c071b06148031c08456804b7181ab13c0aff0a0c2838f58db922","abstract_canon_sha256":"36e036c72aebce8ab8d23f66df3ef3b7aa199ad5663961a1f6bd1c67c89a6509"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"11b5a1ea52a7b3ed049932a7929b6cf5e71b1d6e43c56eed47329a1ed24a9d78","last_reissued_at":"2026-07-31T01:09:27.269360Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:09:27.269360Z"},"graph_snapshot":{"paper":{"title":"Select or Project? Evaluating Lower-dimensional Vectors for LLM Training Data Explanations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Benjamin Roth, Loris Schoenegger, Lukas Hinterleitner","submitted_at":"2026-01-23T11:15:20Z","abstract_excerpt":"Gradient-based methods for instance-based explanation for large language models (LLMs) are hindered by the immense dimensionality of model gradients. In practice, influence estimation is restricted to a subset of model parameters to make computation tractable, but this subset is often chosen ad hoc and rarely justified by systematic evaluation. This paper investigates if it is better to create low-dimensional representations by selecting a small, architecturally informed subset of model components or by projecting the full gradients into a lower-dimensional space. Using a novel benchmark, we s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.16651","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/2601.16651/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":"2601.16651","created_at":"2026-07-31T01:09:27.272509+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.16651v3","created_at":"2026-07-31T01:09:27.272509+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.16651","created_at":"2026-07-31T01:09:27.272509+00:00"},{"alias_kind":"pith_short_12","alias_value":"CG22D2SSU6Z6","created_at":"2026-07-31T01:09:27.272509+00:00"},{"alias_kind":"pith_short_16","alias_value":"CG22D2SSU6Z62BEZ","created_at":"2026-07-31T01:09:27.272509+00:00"},{"alias_kind":"pith_short_8","alias_value":"CG22D2SS","created_at":"2026-07-31T01:09:27.272509+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X","json":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X.json","graph_json":"https://pith.science/api/pith-number/CG22D2SSU6Z62BEZGKTZFG3M6X/graph.json","events_json":"https://pith.science/api/pith-number/CG22D2SSU6Z62BEZGKTZFG3M6X/events.json","paper":"https://pith.science/paper/CG22D2SS"},"agent_actions":{"view_html":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X","download_json":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X.json","view_paper":"https://pith.science/paper/CG22D2SS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.16651&json=true","fetch_graph":"https://pith.science/api/pith-number/CG22D2SSU6Z62BEZGKTZFG3M6X/graph.json","fetch_events":"https://pith.science/api/pith-number/CG22D2SSU6Z62BEZGKTZFG3M6X/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X/action/storage_attestation","attest_author":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X/action/author_attestation","sign_citation":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X/action/citation_signature","submit_replication":"https://pith.science/pith/CG22D2SSU6Z62BEZGKTZFG3M6X/action/replication_record"}},"created_at":"2026-07-31T01:09:27.272509+00:00","updated_at":"2026-07-31T01:09:27.272509+00:00"}