{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FE77OSEYWIOGDW4FHKTG3KSF4R","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2dbba9817a78880b410ab2dfc2bcaaa41c2e795daff0685843aafb3f4043dc98","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-08-12T06:15:33Z","title_canon_sha256":"4a9e6171aa2e0b60a22b01dd771a79ef70057ca5b1ffb6e3ecb9c9aa3cf4a9e7"},"schema_version":"1.0","source":{"id":"2608.11698","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.11698","created_at":"2026-08-13T01:27:52Z"},{"alias_kind":"arxiv_version","alias_value":"2608.11698v1","created_at":"2026-08-13T01:27:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.11698","created_at":"2026-08-13T01:27:52Z"},{"alias_kind":"pith_short_12","alias_value":"FE77OSEYWIOG","created_at":"2026-08-13T01:27:52Z"},{"alias_kind":"pith_short_16","alias_value":"FE77OSEYWIOGDW4F","created_at":"2026-08-13T01:27:52Z"},{"alias_kind":"pith_short_8","alias_value":"FE77OSEY","created_at":"2026-08-13T01:27:52Z"}],"graph_snapshots":[{"event_id":"sha256:3e8f0bc55ae02f447ae051a755f4f622b266c1f2c5388ddff5dfe1efb39cd156","target":"graph","created_at":"2026-08-13T01:27:52Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2608.11698/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"On-policy distillation (OPD) trains a student on its own trajectories under dense token-level supervision from a teacher. Reward-extrapolation methods such as ExOPD amplify the teacher-reference log-likelihood ratio to move beyond direct imitation, but apply a single global coefficient $\\lambda$ to every token. This can drive the student to fit extreme peaks in the implicit reward, causing reward hacking and unstable training, and the optimal $\\lambda$ varies across domains, requiring costly sweeps. We propose REOPD, a reliability-adaptive reward extrapolation framework for OPD. REOPD combines","authors_text":"Guohang Yan, Hanyang Lu, Houyuan Qin, Lichao Ma, Pinlong Cai, Yang Sun, Yao Zhu, Yuxin Liu","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-08-12T06:15:33Z","title":"REOPD: Reliability-Adaptive Reward Extrapolation for On-Policy Distillation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.11698","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:957d960c2e1248390108fce5262509855419b090d19a90c39fcdccab220df13d","target":"record","created_at":"2026-08-13T01:27:52Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2dbba9817a78880b410ab2dfc2bcaaa41c2e795daff0685843aafb3f4043dc98","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-08-12T06:15:33Z","title_canon_sha256":"4a9e6171aa2e0b60a22b01dd771a79ef70057ca5b1ffb6e3ecb9c9aa3cf4a9e7"},"schema_version":"1.0","source":{"id":"2608.11698","kind":"arxiv","version":1}},"canonical_sha256":"293ff74898b21c61db853aa66daa45e4791616170cde446b57306154a199d0db","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"293ff74898b21c61db853aa66daa45e4791616170cde446b57306154a199d0db","first_computed_at":"2026-08-13T01:27:52.309853Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-13T01:27:52.309853Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"S5unvXLCSphULxu3p/FKw/mXHnsZlkdeZ7Qa36hww9WbfrII56rL/xIOdCWox5cG2bau6jnd7/JNDgCHDqbHBA==","signature_status":"signed_v1","signed_at":"2026-08-13T01:27:52.312014Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.11698","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:957d960c2e1248390108fce5262509855419b090d19a90c39fcdccab220df13d","sha256:3e8f0bc55ae02f447ae051a755f4f622b266c1f2c5388ddff5dfe1efb39cd156"],"state_sha256":"d8645c8e8dbee4c9edb6fa209a332855c99072ba0a2ded420fe5faa8c88a1a95"}