{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CQV4OGF4ELRPT4QMJLFFGK57PZ","short_pith_number":"pith:CQV4OGF4","schema_version":"1.0","canonical_sha256":"142bc718bc22e2f9f20c4aca532bbf7e6d591a0900398a389f508cc2f8c4f6e1","source":{"kind":"arxiv","id":"2608.09226","version":1},"attestation_state":"computed","paper":{"title":"RL-Native Distillation: Exploiting Scored Trajectories for Few-Step Image Generation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Bingbing Ni, Fangao Zeng, Hao Zhou, Mengfei Xu, Pipei Huang, Sicong Kang, Wei Li, Yuhan Li","submitted_at":"2026-08-10T07:49:05Z","abstract_excerpt":"Efficient text-to-image generation requires both reinforcement-learning (RL)-based reward alignment and few-step distillation, yet these procedures are typically performed sequentially, increasing training cost and risking the loss of reward gains during compression. We instead take an RL-native perspective: diffusion RL already generates reward-scored finite-step trajectories, whose intermediate states provide a natural source of distillation supervision rather than a disposable byproduct of sampling. Based on this insight, we propose REST (Reward-Enhanced Scored-Trajectory Distillation), a s"},"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":"2608.09226","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-10T07:49:05Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d2691689586f225c41a9cfc9deb9a4509471de5f74bad2861fa41e0164f82b28","abstract_canon_sha256":"1c86ed00b0d0895503e65b510a4280e27f01c7e9ac4300148758846936254da1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T02:22:07.593644Z","signature_b64":"toVQHgoITJPYxsev6W+v29g848LaP/VR7FCftl8djCOxE+w9WDa1nXTZwB/pUEr/TBspQr+VzOzLo9zGqZh1BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"142bc718bc22e2f9f20c4aca532bbf7e6d591a0900398a389f508cc2f8c4f6e1","last_reissued_at":"2026-08-11T02:22:07.592147Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T02:22:07.592147Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RL-Native Distillation: Exploiting Scored Trajectories for Few-Step Image Generation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Bingbing Ni, Fangao Zeng, Hao Zhou, Mengfei Xu, Pipei Huang, Sicong Kang, Wei Li, Yuhan Li","submitted_at":"2026-08-10T07:49:05Z","abstract_excerpt":"Efficient text-to-image generation requires both reinforcement-learning (RL)-based reward alignment and few-step distillation, yet these procedures are typically performed sequentially, increasing training cost and risking the loss of reward gains during compression. We instead take an RL-native perspective: diffusion RL already generates reward-scored finite-step trajectories, whose intermediate states provide a natural source of distillation supervision rather than a disposable byproduct of sampling. Based on this insight, we propose REST (Reward-Enhanced Scored-Trajectory Distillation), a s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.09226","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/2608.09226/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":"2608.09226","created_at":"2026-08-11T02:22:07.592757+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.09226v1","created_at":"2026-08-11T02:22:07.592757+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.09226","created_at":"2026-08-11T02:22:07.592757+00:00"},{"alias_kind":"pith_short_12","alias_value":"CQV4OGF4ELRP","created_at":"2026-08-11T02:22:07.592757+00:00"},{"alias_kind":"pith_short_16","alias_value":"CQV4OGF4ELRPT4QM","created_at":"2026-08-11T02:22:07.592757+00:00"},{"alias_kind":"pith_short_8","alias_value":"CQV4OGF4","created_at":"2026-08-11T02:22:07.592757+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/CQV4OGF4ELRPT4QMJLFFGK57PZ","json":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ.json","graph_json":"https://pith.science/api/pith-number/CQV4OGF4ELRPT4QMJLFFGK57PZ/graph.json","events_json":"https://pith.science/api/pith-number/CQV4OGF4ELRPT4QMJLFFGK57PZ/events.json","paper":"https://pith.science/paper/CQV4OGF4"},"agent_actions":{"view_html":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ","download_json":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ.json","view_paper":"https://pith.science/paper/CQV4OGF4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.09226&json=true","fetch_graph":"https://pith.science/api/pith-number/CQV4OGF4ELRPT4QMJLFFGK57PZ/graph.json","fetch_events":"https://pith.science/api/pith-number/CQV4OGF4ELRPT4QMJLFFGK57PZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ/action/storage_attestation","attest_author":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ/action/author_attestation","sign_citation":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ/action/citation_signature","submit_replication":"https://pith.science/pith/CQV4OGF4ELRPT4QMJLFFGK57PZ/action/replication_record"}},"created_at":"2026-08-11T02:22:07.592757+00:00","updated_at":"2026-08-11T02:22:07.592757+00:00"}