{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:PIZSYZ2R2FVNIF5HZSTPRFXP2N","short_pith_number":"pith:PIZSYZ2R","schema_version":"1.0","canonical_sha256":"7a332c6751d16ad417a7cca6f896efd347b85fce832a746615a6931bad809876","source":{"kind":"arxiv","id":"2010.05358","version":1},"attestation_state":"computed","paper":{"title":"Learning Which Features Matter: RoBERTa Acquires a Preference for Linguistic Generalizations (Eventually)","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alex Warstadt, Haau-sing Li, Haokun Liu, Samuel R. Bowman, Yian Zhang","submitted_at":"2020-10-11T22:09:27Z","abstract_excerpt":"One reason pretraining on self-supervised linguistic tasks is effective is that it teaches models features that are helpful for language understanding. However, we want pretrained models to learn not only to represent linguistic features, but also to use those features preferentially during fine-turning. With this goal in mind, we introduce a new English-language diagnostic set called MSGS (the Mixed Signals Generalization Set), which consists of 20 ambiguous binary classification tasks that we use to test whether a pretrained model prefers linguistic or surface generalizations during fine-tun"},"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":"2010.05358","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-11T22:09:27Z","cross_cats_sorted":[],"title_canon_sha256":"58e510ce25a16c414353057ac89d346af77d7060c80c420e6cedce179c4d182e","abstract_canon_sha256":"ad007e17c3b3068f71ca93b2ec6ca23999f886a23c59b0d39d90daeed29f309b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:42:05.475682Z","signature_b64":"KjjwdjnUgB7YBRB585+dAMgMufPs80fGnX6DcUkD3926YGaUoLOhboXT+j9VUCECAQshRpOE2MQg6cl2m8OtDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7a332c6751d16ad417a7cca6f896efd347b85fce832a746615a6931bad809876","last_reissued_at":"2026-07-05T01:42:05.475272Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:42:05.475272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Which Features Matter: RoBERTa Acquires a Preference for Linguistic Generalizations (Eventually)","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alex Warstadt, Haau-sing Li, Haokun Liu, Samuel R. Bowman, Yian Zhang","submitted_at":"2020-10-11T22:09:27Z","abstract_excerpt":"One reason pretraining on self-supervised linguistic tasks is effective is that it teaches models features that are helpful for language understanding. However, we want pretrained models to learn not only to represent linguistic features, but also to use those features preferentially during fine-turning. With this goal in mind, we introduce a new English-language diagnostic set called MSGS (the Mixed Signals Generalization Set), which consists of 20 ambiguous binary classification tasks that we use to test whether a pretrained model prefers linguistic or surface generalizations during fine-tun"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.05358","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/2010.05358/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":"2010.05358","created_at":"2026-07-05T01:42:05.475347+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.05358v1","created_at":"2026-07-05T01:42:05.475347+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.05358","created_at":"2026-07-05T01:42:05.475347+00:00"},{"alias_kind":"pith_short_12","alias_value":"PIZSYZ2R2FVN","created_at":"2026-07-05T01:42:05.475347+00:00"},{"alias_kind":"pith_short_16","alias_value":"PIZSYZ2R2FVNIF5H","created_at":"2026-07-05T01:42:05.475347+00:00"},{"alias_kind":"pith_short_8","alias_value":"PIZSYZ2R","created_at":"2026-07-05T01:42:05.475347+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02953","citing_title":"Linguistic Productivity in Large Language Models: Models Coerce, but do not Preempt","ref_index":170,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N","json":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N.json","graph_json":"https://pith.science/api/pith-number/PIZSYZ2R2FVNIF5HZSTPRFXP2N/graph.json","events_json":"https://pith.science/api/pith-number/PIZSYZ2R2FVNIF5HZSTPRFXP2N/events.json","paper":"https://pith.science/paper/PIZSYZ2R"},"agent_actions":{"view_html":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N","download_json":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N.json","view_paper":"https://pith.science/paper/PIZSYZ2R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.05358&json=true","fetch_graph":"https://pith.science/api/pith-number/PIZSYZ2R2FVNIF5HZSTPRFXP2N/graph.json","fetch_events":"https://pith.science/api/pith-number/PIZSYZ2R2FVNIF5HZSTPRFXP2N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N/action/storage_attestation","attest_author":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N/action/author_attestation","sign_citation":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N/action/citation_signature","submit_replication":"https://pith.science/pith/PIZSYZ2R2FVNIF5HZSTPRFXP2N/action/replication_record"}},"created_at":"2026-07-05T01:42:05.475347+00:00","updated_at":"2026-07-05T01:42:05.475347+00:00"}