{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:N5IPFUTS7FMRJ4AXXUE27IPLFA","short_pith_number":"pith:N5IPFUTS","schema_version":"1.0","canonical_sha256":"6f50f2d272f95914f017bd09afa1eb280c18a479b14247c5008191c904f9cc91","source":{"kind":"arxiv","id":"2005.00175","version":3},"attestation_state":"computed","paper":{"title":"Selecting Informative Contexts Improves Language Model Finetuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Huth, Javier Turek, Nicole Beckage, Richard Antonello","submitted_at":"2020-05-01T02:01:18Z","abstract_excerpt":"Language model fine-tuning is essential for modern natural language processing, but is computationally expensive and time-consuming. Further, the effectiveness of fine-tuning is limited by the inclusion of training examples that negatively affect performance. Here we present a general fine-tuning method that we call information gain filtration for improving the overall training efficiency and final performance of language model fine-tuning. We define the information gain of an example as the improvement on a test metric after training on that example. A secondary learner is then trained to app"},"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":"2005.00175","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-05-01T02:01:18Z","cross_cats_sorted":[],"title_canon_sha256":"fd41aac96594acf374c3d3a23f087cfe8bdbb01c1d2341a13d241597cf6272c7","abstract_canon_sha256":"b88e437fde1ad4b3622811484e669d5253116fd5f01b13c70833afda53834b00"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:24:53.786614Z","signature_b64":"3ABrW0C5p8GJ9UYwHBUD9w5dsnTlVAo9HjyKoDxNfk5rcZU9frAL+mze52glyyiOHyZKfwC3yn18389XF8lQDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6f50f2d272f95914f017bd09afa1eb280c18a479b14247c5008191c904f9cc91","last_reissued_at":"2026-07-05T04:24:53.785996Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:24:53.785996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Selecting Informative Contexts Improves Language Model Finetuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Huth, Javier Turek, Nicole Beckage, Richard Antonello","submitted_at":"2020-05-01T02:01:18Z","abstract_excerpt":"Language model fine-tuning is essential for modern natural language processing, but is computationally expensive and time-consuming. Further, the effectiveness of fine-tuning is limited by the inclusion of training examples that negatively affect performance. Here we present a general fine-tuning method that we call information gain filtration for improving the overall training efficiency and final performance of language model fine-tuning. We define the information gain of an example as the improvement on a test metric after training on that example. A secondary learner is then trained to app"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.00175","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/2005.00175/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":"2005.00175","created_at":"2026-07-05T04:24:53.786066+00:00"},{"alias_kind":"arxiv_version","alias_value":"2005.00175v3","created_at":"2026-07-05T04:24:53.786066+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.00175","created_at":"2026-07-05T04:24:53.786066+00:00"},{"alias_kind":"pith_short_12","alias_value":"N5IPFUTS7FMR","created_at":"2026-07-05T04:24:53.786066+00:00"},{"alias_kind":"pith_short_16","alias_value":"N5IPFUTS7FMRJ4AX","created_at":"2026-07-05T04:24:53.786066+00:00"},{"alias_kind":"pith_short_8","alias_value":"N5IPFUTS","created_at":"2026-07-05T04:24:53.786066+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.19051","citing_title":"Efficient Data Selection at Scale via Influence Distillation","ref_index":3,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA","json":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA.json","graph_json":"https://pith.science/api/pith-number/N5IPFUTS7FMRJ4AXXUE27IPLFA/graph.json","events_json":"https://pith.science/api/pith-number/N5IPFUTS7FMRJ4AXXUE27IPLFA/events.json","paper":"https://pith.science/paper/N5IPFUTS"},"agent_actions":{"view_html":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA","download_json":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA.json","view_paper":"https://pith.science/paper/N5IPFUTS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2005.00175&json=true","fetch_graph":"https://pith.science/api/pith-number/N5IPFUTS7FMRJ4AXXUE27IPLFA/graph.json","fetch_events":"https://pith.science/api/pith-number/N5IPFUTS7FMRJ4AXXUE27IPLFA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA/action/storage_attestation","attest_author":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA/action/author_attestation","sign_citation":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA/action/citation_signature","submit_replication":"https://pith.science/pith/N5IPFUTS7FMRJ4AXXUE27IPLFA/action/replication_record"}},"created_at":"2026-07-05T04:24:53.786066+00:00","updated_at":"2026-07-05T04:24:53.786066+00:00"}