{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:2HMPTTBFE6JMP7S35MN4TGOBYQ","short_pith_number":"pith:2HMPTTBF","canonical_record":{"source":{"id":"2605.25582","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T08:32:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2fb1c136ef6e2a3caa39088ca1bb923e46efb440a92ecfca0cdcd205514ebf4b","abstract_canon_sha256":"c7e59d3074ded757df28b6a233df98d071785753c223191f662a0136f86bd435"},"schema_version":"1.0"},"canonical_sha256":"d1d8f9cc252792c7fe5beb1bc999c1c43bb0323483616b6941e0230727fcb895","source":{"kind":"arxiv","id":"2605.25582","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.25582","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"arxiv_version","alias_value":"2605.25582v1","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25582","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_12","alias_value":"2HMPTTBFE6JM","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_16","alias_value":"2HMPTTBFE6JMP7S3","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_8","alias_value":"2HMPTTBF","created_at":"2026-05-26T02:04:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:2HMPTTBFE6JMP7S35MN4TGOBYQ","target":"record","payload":{"canonical_record":{"source":{"id":"2605.25582","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T08:32:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2fb1c136ef6e2a3caa39088ca1bb923e46efb440a92ecfca0cdcd205514ebf4b","abstract_canon_sha256":"c7e59d3074ded757df28b6a233df98d071785753c223191f662a0136f86bd435"},"schema_version":"1.0"},"canonical_sha256":"d1d8f9cc252792c7fe5beb1bc999c1c43bb0323483616b6941e0230727fcb895","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:44.184832Z","signature_b64":"P/ISjDm7ZnQ95kDLo0Y/ltIGfN9inz//9NI2YPeZRgpb5wHF7iKcSKjqOJWVisgTUDAma6VsAuwgNwT9TYNvAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d1d8f9cc252792c7fe5beb1bc999c1c43bb0323483616b6941e0230727fcb895","last_reissued_at":"2026-05-26T02:04:44.184186Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:44.184186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.25582","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":"2026-05-26T02:04:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fN0l3O2i6EKJmE3alvsPIpF+OzC6hJhE+qCunNF5L0WOOdbt9hcwUyAF/pflk1OOYvZlc5ouWyRhm3TkC0qTDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T08:33:08.204962Z"},"content_sha256":"5adfc8bfdb046b6fdd0257b7fecf4ee48b89332e54cc46db00c2d4713c6e2566","schema_version":"1.0","event_id":"sha256:5adfc8bfdb046b6fdd0257b7fecf4ee48b89332e54cc46db00c2d4713c6e2566"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:2HMPTTBFE6JMP7S35MN4TGOBYQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Extreme Region Policy Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Changyu Chen, Rui Yan, Xiting Wang","submitted_at":"2026-05-25T08:32:24Z","abstract_excerpt":"Reinforcement learning for large language models faces a fundamental trade-off between sample efficiency and asymptotic performance: strictly on-policy methods discard trajectories after a single update, while off-policy reuse introduces distribution mismatch that existing trust-region techniques mitigate primarily by enforcing conservative optimization, often leaving rich training signals underutilized. To investigate this, we perform extensive off-policy updates on fixed data. Our experiments reveal that aggressive multi-step optimization brings rapid initial gains, but excessive updates cau"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25582","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/2605.25582/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":"2026-05-26T02:04:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XvZGqMK7iDIAzAJKdGI1MDN9+/BwdZ3XasrPFABmB5QXywU5yuVvcug+pKwb7hJiWKyG1mli7/4ZF0t3V0r0DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T08:33:08.205622Z"},"content_sha256":"ad0318fddf394fe9490e831defc02f1fee0431ee85f455d1a2c259442180a517","schema_version":"1.0","event_id":"sha256:ad0318fddf394fe9490e831defc02f1fee0431ee85f455d1a2c259442180a517"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/bundle.json","state_url":"https://pith.science/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/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":"2026-05-28T08:33:08Z","links":{"resolver":"https://pith.science/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ","bundle":"https://pith.science/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/bundle.json","state":"https://pith.science/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2HMPTTBFE6JMP7S35MN4TGOBYQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:2HMPTTBFE6JMP7S35MN4TGOBYQ","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":"c7e59d3074ded757df28b6a233df98d071785753c223191f662a0136f86bd435","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T08:32:24Z","title_canon_sha256":"2fb1c136ef6e2a3caa39088ca1bb923e46efb440a92ecfca0cdcd205514ebf4b"},"schema_version":"1.0","source":{"id":"2605.25582","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.25582","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"arxiv_version","alias_value":"2605.25582v1","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25582","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_12","alias_value":"2HMPTTBFE6JM","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_16","alias_value":"2HMPTTBFE6JMP7S3","created_at":"2026-05-26T02:04:44Z"},{"alias_kind":"pith_short_8","alias_value":"2HMPTTBF","created_at":"2026-05-26T02:04:44Z"}],"graph_snapshots":[{"event_id":"sha256:ad0318fddf394fe9490e831defc02f1fee0431ee85f455d1a2c259442180a517","target":"graph","created_at":"2026-05-26T02:04:44Z","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/2605.25582/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning for large language models faces a fundamental trade-off between sample efficiency and asymptotic performance: strictly on-policy methods discard trajectories after a single update, while off-policy reuse introduces distribution mismatch that existing trust-region techniques mitigate primarily by enforcing conservative optimization, often leaving rich training signals underutilized. To investigate this, we perform extensive off-policy updates on fixed data. Our experiments reveal that aggressive multi-step optimization brings rapid initial gains, but excessive updates cau","authors_text":"Changyu Chen, Rui Yan, Xiting Wang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T08:32:24Z","title":"Extreme Region Policy Distillation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25582","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":"sha256:5adfc8bfdb046b6fdd0257b7fecf4ee48b89332e54cc46db00c2d4713c6e2566","target":"record","created_at":"2026-05-26T02:04:44Z","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":"c7e59d3074ded757df28b6a233df98d071785753c223191f662a0136f86bd435","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T08:32:24Z","title_canon_sha256":"2fb1c136ef6e2a3caa39088ca1bb923e46efb440a92ecfca0cdcd205514ebf4b"},"schema_version":"1.0","source":{"id":"2605.25582","kind":"arxiv","version":1}},"canonical_sha256":"d1d8f9cc252792c7fe5beb1bc999c1c43bb0323483616b6941e0230727fcb895","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d1d8f9cc252792c7fe5beb1bc999c1c43bb0323483616b6941e0230727fcb895","first_computed_at":"2026-05-26T02:04:44.184186Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T02:04:44.184186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"P/ISjDm7ZnQ95kDLo0Y/ltIGfN9inz//9NI2YPeZRgpb5wHF7iKcSKjqOJWVisgTUDAma6VsAuwgNwT9TYNvAg==","signature_status":"signed_v1","signed_at":"2026-05-26T02:04:44.184832Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.25582","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5adfc8bfdb046b6fdd0257b7fecf4ee48b89332e54cc46db00c2d4713c6e2566","sha256:ad0318fddf394fe9490e831defc02f1fee0431ee85f455d1a2c259442180a517"],"state_sha256":"216f37ccd661bf2624e7dabc3406cb405da845679dbb154bff88995c4231355f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OuQCN3ETMzoJ5ToSWs365QY7ZAkHLLJTDKr+9Qzg8PFIsAFti3F/hK7mA1bfFvk04ZMALWnsbksUcMaK40PJDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T08:33:08.208588Z","bundle_sha256":"d48cb9f60f7b9a291d030253d580c667c3632638d447b3d0393acada024afbe4"}}