{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:VDHUKXMOOBH3XAHT7IYDSMLX4F","short_pith_number":"pith:VDHUKXMO","schema_version":"1.0","canonical_sha256":"a8cf455d8e704fbb80f3fa30393177e15ab8dd7aff44e079b8b9925214c64d5e","source":{"kind":"arxiv","id":"1910.07117","version":5},"attestation_state":"computed","paper":{"title":"Analyzing the Forgetting Problem in the Pretrain-Finetuning of Dialogue Response Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Bing Liu, Fuchun Peng, James Glass, Jun Liu, Kyunghyun Cho, Myle Ott, Tianxing He","submitted_at":"2019-10-16T01:10:10Z","abstract_excerpt":"In this work, we study how the finetuning stage in the pretrain-finetune framework changes the behavior of a pretrained neural language generator. We focus on the transformer encoder-decoder model for the open-domain dialogue response generation task. Our major finding is that after standard finetuning, the model forgets some of the important language generation skills acquired during large-scale pretraining. We demonstrate the forgetting phenomenon through a set of detailed behavior analysis from the perspectives of knowledge transfer, context sensitivity, and function space projection. As a "},"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":"1910.07117","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-10-16T01:10:10Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"32fa9e38dc1e58067325101ef8df2ea00c52d406cb23e2d5a1cd04687055275d","abstract_canon_sha256":"6775b46922a426a673ac0b4b64257770c29573b02003ca207614189be4fd6e15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:07:13.358149Z","signature_b64":"mH3fyenomvM5IAMycIk2Y0v+9pGgXwGm8PGIscp/bez4l7pcH/NDtrEAJ+QtGCM76IyH5knJ3cIaEAG9/UeuCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a8cf455d8e704fbb80f3fa30393177e15ab8dd7aff44e079b8b9925214c64d5e","last_reissued_at":"2026-07-05T02:07:13.357590Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:07:13.357590Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analyzing the Forgetting Problem in the Pretrain-Finetuning of Dialogue Response Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Bing Liu, Fuchun Peng, James Glass, Jun Liu, Kyunghyun Cho, Myle Ott, Tianxing He","submitted_at":"2019-10-16T01:10:10Z","abstract_excerpt":"In this work, we study how the finetuning stage in the pretrain-finetune framework changes the behavior of a pretrained neural language generator. We focus on the transformer encoder-decoder model for the open-domain dialogue response generation task. Our major finding is that after standard finetuning, the model forgets some of the important language generation skills acquired during large-scale pretraining. We demonstrate the forgetting phenomenon through a set of detailed behavior analysis from the perspectives of knowledge transfer, context sensitivity, and function space projection. As a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.07117","kind":"arxiv","version":5},"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/1910.07117/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":"1910.07117","created_at":"2026-07-05T02:07:13.357659+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.07117v5","created_at":"2026-07-05T02:07:13.357659+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.07117","created_at":"2026-07-05T02:07:13.357659+00:00"},{"alias_kind":"pith_short_12","alias_value":"VDHUKXMOOBH3","created_at":"2026-07-05T02:07:13.357659+00:00"},{"alias_kind":"pith_short_16","alias_value":"VDHUKXMOOBH3XAHT","created_at":"2026-07-05T02:07:13.357659+00:00"},{"alias_kind":"pith_short_8","alias_value":"VDHUKXMO","created_at":"2026-07-05T02:07:13.357659+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/VDHUKXMOOBH3XAHT7IYDSMLX4F","json":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F.json","graph_json":"https://pith.science/api/pith-number/VDHUKXMOOBH3XAHT7IYDSMLX4F/graph.json","events_json":"https://pith.science/api/pith-number/VDHUKXMOOBH3XAHT7IYDSMLX4F/events.json","paper":"https://pith.science/paper/VDHUKXMO"},"agent_actions":{"view_html":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F","download_json":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F.json","view_paper":"https://pith.science/paper/VDHUKXMO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.07117&json=true","fetch_graph":"https://pith.science/api/pith-number/VDHUKXMOOBH3XAHT7IYDSMLX4F/graph.json","fetch_events":"https://pith.science/api/pith-number/VDHUKXMOOBH3XAHT7IYDSMLX4F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F/action/storage_attestation","attest_author":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F/action/author_attestation","sign_citation":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F/action/citation_signature","submit_replication":"https://pith.science/pith/VDHUKXMOOBH3XAHT7IYDSMLX4F/action/replication_record"}},"created_at":"2026-07-05T02:07:13.357659+00:00","updated_at":"2026-07-05T02:07:13.357659+00:00"}