{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:BLKURIMSDBKKIY54PBEMNMXKTX","short_pith_number":"pith:BLKURIMS","schema_version":"1.0","canonical_sha256":"0ad548a1921854a463bc7848c6b2ea9dffa8bb2e67656128e0d055f91f652e5f","source":{"kind":"arxiv","id":"2211.16773","version":5},"attestation_state":"computed","paper":{"title":"KRLS: Improving End-to-End Response Generation in Task Oriented Dialog with Reinforced Keywords Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kun Qian, Qingyang Wu, Xiao Yu, Zhou Yu","submitted_at":"2022-11-30T06:27:46Z","abstract_excerpt":"In task-oriented dialogs (TOD), reinforcement learning (RL) algorithms train a model to directly optimize response for task-related metrics. However, RL needs to perform exploration, which can be time-consuming due to the slow auto-regressive sequence generation process. We investigate an approach to create a more efficient RL-based algorithm to improve TOD performance in an offline setting. First, we use a faster generation procedure that samples from independent next-word distributions after training the language model (LM) with supervised learning. We then introduce a fine-grained reward fu"},"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":"2211.16773","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-11-30T06:27:46Z","cross_cats_sorted":[],"title_canon_sha256":"4ded85d4a11b72e20f174c629b1ea093463aa63fe046abb8d3633ba99823c773","abstract_canon_sha256":"780118b4df8b6e3b68e4fa21755edc59148c5b1d82f72f672a8ffb2fc0dc6faf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:02:47.792583Z","signature_b64":"twDtXZpicCXUMbM6at8DdK9qHnrFOLnpXjX4rp+4kyuU5khnAYiDpuiojGy8LWNXl7Do0xWNcqRJ6iHCweHDDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ad548a1921854a463bc7848c6b2ea9dffa8bb2e67656128e0d055f91f652e5f","last_reissued_at":"2026-07-05T07:02:47.791997Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:02:47.791997Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"KRLS: Improving End-to-End Response Generation in Task Oriented Dialog with Reinforced Keywords Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Kun Qian, Qingyang Wu, Xiao Yu, Zhou Yu","submitted_at":"2022-11-30T06:27:46Z","abstract_excerpt":"In task-oriented dialogs (TOD), reinforcement learning (RL) algorithms train a model to directly optimize response for task-related metrics. However, RL needs to perform exploration, which can be time-consuming due to the slow auto-regressive sequence generation process. We investigate an approach to create a more efficient RL-based algorithm to improve TOD performance in an offline setting. First, we use a faster generation procedure that samples from independent next-word distributions after training the language model (LM) with supervised learning. We then introduce a fine-grained reward fu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.16773","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/2211.16773/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":"2211.16773","created_at":"2026-07-05T07:02:47.792058+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.16773v5","created_at":"2026-07-05T07:02:47.792058+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.16773","created_at":"2026-07-05T07:02:47.792058+00:00"},{"alias_kind":"pith_short_12","alias_value":"BLKURIMSDBKK","created_at":"2026-07-05T07:02:47.792058+00:00"},{"alias_kind":"pith_short_16","alias_value":"BLKURIMSDBKKIY54","created_at":"2026-07-05T07:02:47.792058+00:00"},{"alias_kind":"pith_short_8","alias_value":"BLKURIMS","created_at":"2026-07-05T07:02:47.792058+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/BLKURIMSDBKKIY54PBEMNMXKTX","json":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX.json","graph_json":"https://pith.science/api/pith-number/BLKURIMSDBKKIY54PBEMNMXKTX/graph.json","events_json":"https://pith.science/api/pith-number/BLKURIMSDBKKIY54PBEMNMXKTX/events.json","paper":"https://pith.science/paper/BLKURIMS"},"agent_actions":{"view_html":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX","download_json":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX.json","view_paper":"https://pith.science/paper/BLKURIMS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.16773&json=true","fetch_graph":"https://pith.science/api/pith-number/BLKURIMSDBKKIY54PBEMNMXKTX/graph.json","fetch_events":"https://pith.science/api/pith-number/BLKURIMSDBKKIY54PBEMNMXKTX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX/action/storage_attestation","attest_author":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX/action/author_attestation","sign_citation":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX/action/citation_signature","submit_replication":"https://pith.science/pith/BLKURIMSDBKKIY54PBEMNMXKTX/action/replication_record"}},"created_at":"2026-07-05T07:02:47.792058+00:00","updated_at":"2026-07-05T07:02:47.792058+00:00"}