{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RL5QHOM3LWOFMBOAVMGKJHHQMU","short_pith_number":"pith:RL5QHOM3","canonical_record":{"source":{"id":"2607.05394","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-06T17:59:58Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"7bfaf50a8826860356d1966ed9c0389f1660a976504b9023b668601f66e4db1e","abstract_canon_sha256":"1b44169fd68d18ab9173fbb003cbfc80f7cba93cc7bc36f2f218dc5c89ae3d95"},"schema_version":"1.0"},"canonical_sha256":"8afb03b99b5d9c5605c0ab0ca49cf065046555ab89bbe2ffc6ee4fe901572e7f","source":{"kind":"arxiv","id":"2607.05394","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.05394","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"arxiv_version","alias_value":"2607.05394v1","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.05394","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_12","alias_value":"RL5QHOM3LWOF","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_16","alias_value":"RL5QHOM3LWOFMBOA","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_8","alias_value":"RL5QHOM3","created_at":"2026-07-07T03:19:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RL5QHOM3LWOFMBOAVMGKJHHQMU","target":"record","payload":{"canonical_record":{"source":{"id":"2607.05394","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-06T17:59:58Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"7bfaf50a8826860356d1966ed9c0389f1660a976504b9023b668601f66e4db1e","abstract_canon_sha256":"1b44169fd68d18ab9173fbb003cbfc80f7cba93cc7bc36f2f218dc5c89ae3d95"},"schema_version":"1.0"},"canonical_sha256":"8afb03b99b5d9c5605c0ab0ca49cf065046555ab89bbe2ffc6ee4fe901572e7f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T03:19:29.585037Z","signature_b64":"R2CXQlHC5nbNzTcgjxWZC1C2U5dZgsIbumyt4q5hQO8mcFARU79VGjdRwWUPc2lMUSLX6P3LRrDl7R703C0ACg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8afb03b99b5d9c5605c0ab0ca49cf065046555ab89bbe2ffc6ee4fe901572e7f","last_reissued_at":"2026-07-07T03:19:29.584619Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T03:19:29.584619Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.05394","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-07T03:19:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BlZBEV8xUX+YvlgQVwTG1iNdYL7bAt2GPSA+l44mbpAneYEwFo4KNQ33ooSFaSX2QvRC/4sk8V+S7Nhf7M58Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-12T08:19:31.046306Z"},"content_sha256":"e0525205260c93232e9cc4e1166b85e2f4b347efc33e7860ab7bdb228f72449d","schema_version":"1.0","event_id":"sha256:e0525205260c93232e9cc4e1166b85e2f4b347efc33e7860ab7bdb228f72449d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RL5QHOM3LWOFMBOAVMGKJHHQMU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Weak-to-Strong Generalization via Direct On-Policy Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Bingxiang He, Hanlin Wu, Haohan Chi, Hao Zhou, Huan-ang Gao, Shiyuan Feng, Wei-Ying Ma, Ya-Qin Zhang, Zheng Jiang, Zhilong Zhang","submitted_at":"2026-07-06T17:59:58Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) is a powerful recipe for improving language-model reasoning, but it is expensive to repeat on every new strong model because the target model must generate many rollouts during training. As models scale, post-training itself becomes a bottleneck. We study a weak-to-strong alternative: run RL on a smaller model where rollouts are cheaper, then reuse what that RL run learned to improve a stronger target model. Directly distilling the post-RL weak teacher is not enough, because the teacher's final policy mixes useful RL gains with the limitati"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.05394","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/2607.05394/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-07T03:19:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NMIfOopH/DqypkHvgR9NURdLWsXTLVUUJkBamqp30sKFahBatB0L9f6nSkurW/9qQHFx5gBHu1Xe6Tb97e7hBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-12T08:19:31.047182Z"},"content_sha256":"ca8d9d7dd8371761db3cb902db4e7f000b39754900971f4d4e1b1901ccdebd7f","schema_version":"1.0","event_id":"sha256:ca8d9d7dd8371761db3cb902db4e7f000b39754900971f4d4e1b1901ccdebd7f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/bundle.json","state_url":"https://pith.science/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-12T08:19:31Z","links":{"resolver":"https://pith.science/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU","bundle":"https://pith.science/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/bundle.json","state":"https://pith.science/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RL5QHOM3LWOFMBOAVMGKJHHQMU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RL5QHOM3LWOFMBOAVMGKJHHQMU","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":"1b44169fd68d18ab9173fbb003cbfc80f7cba93cc7bc36f2f218dc5c89ae3d95","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-06T17:59:58Z","title_canon_sha256":"7bfaf50a8826860356d1966ed9c0389f1660a976504b9023b668601f66e4db1e"},"schema_version":"1.0","source":{"id":"2607.05394","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.05394","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"arxiv_version","alias_value":"2607.05394v1","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.05394","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_12","alias_value":"RL5QHOM3LWOF","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_16","alias_value":"RL5QHOM3LWOFMBOA","created_at":"2026-07-07T03:19:29Z"},{"alias_kind":"pith_short_8","alias_value":"RL5QHOM3","created_at":"2026-07-07T03:19:29Z"}],"graph_snapshots":[{"event_id":"sha256:ca8d9d7dd8371761db3cb902db4e7f000b39754900971f4d4e1b1901ccdebd7f","target":"graph","created_at":"2026-07-07T03:19:29Z","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/2607.05394/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) is a powerful recipe for improving language-model reasoning, but it is expensive to repeat on every new strong model because the target model must generate many rollouts during training. As models scale, post-training itself becomes a bottleneck. We study a weak-to-strong alternative: run RL on a smaller model where rollouts are cheaper, then reuse what that RL run learned to improve a stronger target model. Directly distilling the post-RL weak teacher is not enough, because the teacher's final policy mixes useful RL gains with the limitati","authors_text":"Bingxiang He, Hanlin Wu, Haohan Chi, Hao Zhou, Huan-ang Gao, Shiyuan Feng, Wei-Ying Ma, Ya-Qin Zhang, Zheng Jiang, Zhilong Zhang","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-06T17:59:58Z","title":"Weak-to-Strong Generalization via Direct On-Policy Distillation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.05394","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:e0525205260c93232e9cc4e1166b85e2f4b347efc33e7860ab7bdb228f72449d","target":"record","created_at":"2026-07-07T03:19:29Z","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":"1b44169fd68d18ab9173fbb003cbfc80f7cba93cc7bc36f2f218dc5c89ae3d95","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-06T17:59:58Z","title_canon_sha256":"7bfaf50a8826860356d1966ed9c0389f1660a976504b9023b668601f66e4db1e"},"schema_version":"1.0","source":{"id":"2607.05394","kind":"arxiv","version":1}},"canonical_sha256":"8afb03b99b5d9c5605c0ab0ca49cf065046555ab89bbe2ffc6ee4fe901572e7f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8afb03b99b5d9c5605c0ab0ca49cf065046555ab89bbe2ffc6ee4fe901572e7f","first_computed_at":"2026-07-07T03:19:29.584619Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-07T03:19:29.584619Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"R2CXQlHC5nbNzTcgjxWZC1C2U5dZgsIbumyt4q5hQO8mcFARU79VGjdRwWUPc2lMUSLX6P3LRrDl7R703C0ACg==","signature_status":"signed_v1","signed_at":"2026-07-07T03:19:29.585037Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.05394","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e0525205260c93232e9cc4e1166b85e2f4b347efc33e7860ab7bdb228f72449d","sha256:ca8d9d7dd8371761db3cb902db4e7f000b39754900971f4d4e1b1901ccdebd7f"],"state_sha256":"2f5ba410055f4ae456efab4fa7d25a124bf5e7ca3831ff24b9e5d4fe5cc27efd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"T0118m0ZtChIlXbQqdyUm2Qfdf3LpSdkyRjjSbYi2Ulx1pnFazaTQWPFHSLCb10xY3sLXcbTSwUCAiu9SdrUDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-12T08:19:31.050148Z","bundle_sha256":"57a6f600118b9498294d4c775506661535d933dc40df2d2659369dcf7d2fd7fb"}}