{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:6NPUTDP3VARKRPPMQCI2C6NNH4","short_pith_number":"pith:6NPUTDP3","schema_version":"1.0","canonical_sha256":"f35f498dfba822a8bdec8091a179ad3f1f1f9e5b66c0ea9783076777d1ad05dd","source":{"kind":"arxiv","id":"2104.07078","version":1},"attestation_state":"computed","paper":{"title":"UDALM: Unsupervised Domain Adaptation through Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexandros Potamianos, Constantinos Karouzos, Georgios Paraskevopoulos","submitted_at":"2021-04-14T19:05:01Z","abstract_excerpt":"In this work we explore Unsupervised Domain Adaptation (UDA) of pretrained language models for downstream tasks. We introduce UDALM, a fine-tuning procedure, using a mixed classification and Masked Language Model loss, that can adapt to the target domain distribution in a robust and sample efficient manner. Our experiments show that performance of models trained with the mixed loss scales with the amount of available target data and the mixed loss can be effectively used as a stopping criterion during UDA training. Furthermore, we discuss the relationship between A-distance and the target erro"},"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":"2104.07078","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-04-14T19:05:01Z","cross_cats_sorted":[],"title_canon_sha256":"e766deab70305f37f8ba3ce17087987cd43cbf498b4825c616a375573a24e987","abstract_canon_sha256":"edfdfa0c1db300ffc3828416bc7fc82b1cf27d2b15a6d7a4abca2dca4f3bf6f9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:32:17.559101Z","signature_b64":"xi6Ld09QfGR97EOcV8Ao1q8Dsz0RJJp98SA4bQ+P4X+5pps5WOWLLi1ucCZKmhB3xsYnmnAkkRPFox+lKbxSCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f35f498dfba822a8bdec8091a179ad3f1f1f9e5b66c0ea9783076777d1ad05dd","last_reissued_at":"2026-07-05T02:32:17.558275Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:32:17.558275Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UDALM: Unsupervised Domain Adaptation through Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexandros Potamianos, Constantinos Karouzos, Georgios Paraskevopoulos","submitted_at":"2021-04-14T19:05:01Z","abstract_excerpt":"In this work we explore Unsupervised Domain Adaptation (UDA) of pretrained language models for downstream tasks. We introduce UDALM, a fine-tuning procedure, using a mixed classification and Masked Language Model loss, that can adapt to the target domain distribution in a robust and sample efficient manner. Our experiments show that performance of models trained with the mixed loss scales with the amount of available target data and the mixed loss can be effectively used as a stopping criterion during UDA training. Furthermore, we discuss the relationship between A-distance and the target erro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.07078","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/2104.07078/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":"2104.07078","created_at":"2026-07-05T02:32:17.558339+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.07078v1","created_at":"2026-07-05T02:32:17.558339+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.07078","created_at":"2026-07-05T02:32:17.558339+00:00"},{"alias_kind":"pith_short_12","alias_value":"6NPUTDP3VARK","created_at":"2026-07-05T02:32:17.558339+00:00"},{"alias_kind":"pith_short_16","alias_value":"6NPUTDP3VARKRPPM","created_at":"2026-07-05T02:32:17.558339+00:00"},{"alias_kind":"pith_short_8","alias_value":"6NPUTDP3","created_at":"2026-07-05T02:32:17.558339+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.00911","citing_title":"Aligning LLMs with Domain Invariant Reward Models","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4","json":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4.json","graph_json":"https://pith.science/api/pith-number/6NPUTDP3VARKRPPMQCI2C6NNH4/graph.json","events_json":"https://pith.science/api/pith-number/6NPUTDP3VARKRPPMQCI2C6NNH4/events.json","paper":"https://pith.science/paper/6NPUTDP3"},"agent_actions":{"view_html":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4","download_json":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4.json","view_paper":"https://pith.science/paper/6NPUTDP3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.07078&json=true","fetch_graph":"https://pith.science/api/pith-number/6NPUTDP3VARKRPPMQCI2C6NNH4/graph.json","fetch_events":"https://pith.science/api/pith-number/6NPUTDP3VARKRPPMQCI2C6NNH4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4/action/storage_attestation","attest_author":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4/action/author_attestation","sign_citation":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4/action/citation_signature","submit_replication":"https://pith.science/pith/6NPUTDP3VARKRPPMQCI2C6NNH4/action/replication_record"}},"created_at":"2026-07-05T02:32:17.558339+00:00","updated_at":"2026-07-05T02:32:17.558339+00:00"}