{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:OVD5QXTFSXDVKMJ5MVB5P7P6JW","short_pith_number":"pith:OVD5QXTF","canonical_record":{"source":{"id":"2509.02737","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-02T18:33:11Z","cross_cats_sorted":[],"title_canon_sha256":"601a31b5aee022bb41996d8a53159cb11e4af93e2cbd9816b49524c4120811cd","abstract_canon_sha256":"2ac4b3301f9984ee1d8b48de3a78124c5e7c62a055159c82e2a10cddaf7adf05"},"schema_version":"1.0"},"canonical_sha256":"7547d85e6595c755313d6543d7fdfe4db0605b9f8effce6d2891b71983d6e06f","source":{"kind":"arxiv","id":"2509.02737","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.02737","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"arxiv_version","alias_value":"2509.02737v1","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.02737","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_12","alias_value":"OVD5QXTFSXDV","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_16","alias_value":"OVD5QXTFSXDVKMJ5","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_8","alias_value":"OVD5QXTF","created_at":"2026-07-05T12:03:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:OVD5QXTFSXDVKMJ5MVB5P7P6JW","target":"record","payload":{"canonical_record":{"source":{"id":"2509.02737","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-02T18:33:11Z","cross_cats_sorted":[],"title_canon_sha256":"601a31b5aee022bb41996d8a53159cb11e4af93e2cbd9816b49524c4120811cd","abstract_canon_sha256":"2ac4b3301f9984ee1d8b48de3a78124c5e7c62a055159c82e2a10cddaf7adf05"},"schema_version":"1.0"},"canonical_sha256":"7547d85e6595c755313d6543d7fdfe4db0605b9f8effce6d2891b71983d6e06f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:56.789582Z","signature_b64":"xEyO2tIbi0O4D9VnXr3CKpcPUqqIeAkNnggon8kKDGB+Y1a45jvhCnwze1lm4TS+f6RvwRfA60zZW2j0dZM7Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7547d85e6595c755313d6543d7fdfe4db0605b9f8effce6d2891b71983d6e06f","last_reissued_at":"2026-07-05T12:03:56.789065Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:56.789065Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2509.02737","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":"2026-07-05T12:03:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gDO5uKiAhXGmbNSJ1cYhkfu5ZkX/Oe3u/aQ2o7/jp2CtCw8v52bVJM8lk4Pcnfd+ubLT/hRMZt7OQhB2lwdYDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T17:13:06.203136Z"},"content_sha256":"6406d23ef1eeb2e55c8740cfddc02f002cb4c3c75972d86981d3341b912f01f1","schema_version":"1.0","event_id":"sha256:6406d23ef1eeb2e55c8740cfddc02f002cb4c3c75972d86981d3341b912f01f1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:OVD5QXTFSXDVKMJ5MVB5P7P6JW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Imitate Optimal Policy: Prevail and Induce Action Collapse in Policy Gradient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ben Athiwaratkun, Bernard Ghanem, Fengxiang Bie, Haojun Xia, Robert Wu, Shuaiwen Leon Song, Xiaoxia Wu, Yibo Yang, Zhongzhu Zhou, Ziyan Chen","submitted_at":"2025-09-02T18:33:11Z","abstract_excerpt":"Policy gradient (PG) methods in reinforcement learning frequently utilize deep neural networks (DNNs) to learn a shared backbone of feature representations used to compute likelihoods in an action selection layer. Numerous studies have been conducted on the convergence and global optima of policy networks, but few have analyzed representational structures of those underlying networks. While training an optimal policy DNN, we observed that under certain constraints, a gentle structure resembling neural collapse, which we refer to as Action Collapse (AC), emerges. This suggests that 1) the state"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.02737","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/2509.02737/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":"2026-07-05T12:03:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2SvuAaXqeuL/VhtCkLWhTORTAYomx06DV4spkAHW03+0YDfkKP/NOsCxiQ7F4TKTBBU9r3VeTj3lesYzsnAWDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T17:13:06.203646Z"},"content_sha256":"97cc70f08984b39547c02d90f481ddb2e8ea89dd2fbea498af990c5238ebdfe7","schema_version":"1.0","event_id":"sha256:97cc70f08984b39547c02d90f481ddb2e8ea89dd2fbea498af990c5238ebdfe7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/bundle.json","state_url":"https://pith.science/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/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":"2026-08-17T17:13:06Z","links":{"resolver":"https://pith.science/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW","bundle":"https://pith.science/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/bundle.json","state":"https://pith.science/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OVD5QXTFSXDVKMJ5MVB5P7P6JW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:OVD5QXTFSXDVKMJ5MVB5P7P6JW","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":"2ac4b3301f9984ee1d8b48de3a78124c5e7c62a055159c82e2a10cddaf7adf05","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-02T18:33:11Z","title_canon_sha256":"601a31b5aee022bb41996d8a53159cb11e4af93e2cbd9816b49524c4120811cd"},"schema_version":"1.0","source":{"id":"2509.02737","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.02737","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"arxiv_version","alias_value":"2509.02737v1","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.02737","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_12","alias_value":"OVD5QXTFSXDV","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_16","alias_value":"OVD5QXTFSXDVKMJ5","created_at":"2026-07-05T12:03:56Z"},{"alias_kind":"pith_short_8","alias_value":"OVD5QXTF","created_at":"2026-07-05T12:03:56Z"}],"graph_snapshots":[{"event_id":"sha256:97cc70f08984b39547c02d90f481ddb2e8ea89dd2fbea498af990c5238ebdfe7","target":"graph","created_at":"2026-07-05T12:03:56Z","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/2509.02737/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Policy gradient (PG) methods in reinforcement learning frequently utilize deep neural networks (DNNs) to learn a shared backbone of feature representations used to compute likelihoods in an action selection layer. Numerous studies have been conducted on the convergence and global optima of policy networks, but few have analyzed representational structures of those underlying networks. While training an optimal policy DNN, we observed that under certain constraints, a gentle structure resembling neural collapse, which we refer to as Action Collapse (AC), emerges. This suggests that 1) the state","authors_text":"Ben Athiwaratkun, Bernard Ghanem, Fengxiang Bie, Haojun Xia, Robert Wu, Shuaiwen Leon Song, Xiaoxia Wu, Yibo Yang, Zhongzhu Zhou, Ziyan Chen","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-02T18:33:11Z","title":"Imitate Optimal Policy: Prevail and Induce Action Collapse in Policy Gradient"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.02737","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":"sha256:6406d23ef1eeb2e55c8740cfddc02f002cb4c3c75972d86981d3341b912f01f1","target":"record","created_at":"2026-07-05T12:03:56Z","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":"2ac4b3301f9984ee1d8b48de3a78124c5e7c62a055159c82e2a10cddaf7adf05","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-02T18:33:11Z","title_canon_sha256":"601a31b5aee022bb41996d8a53159cb11e4af93e2cbd9816b49524c4120811cd"},"schema_version":"1.0","source":{"id":"2509.02737","kind":"arxiv","version":1}},"canonical_sha256":"7547d85e6595c755313d6543d7fdfe4db0605b9f8effce6d2891b71983d6e06f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7547d85e6595c755313d6543d7fdfe4db0605b9f8effce6d2891b71983d6e06f","first_computed_at":"2026-07-05T12:03:56.789065Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:03:56.789065Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xEyO2tIbi0O4D9VnXr3CKpcPUqqIeAkNnggon8kKDGB+Y1a45jvhCnwze1lm4TS+f6RvwRfA60zZW2j0dZM7Bw==","signature_status":"signed_v1","signed_at":"2026-07-05T12:03:56.789582Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.02737","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6406d23ef1eeb2e55c8740cfddc02f002cb4c3c75972d86981d3341b912f01f1","sha256:97cc70f08984b39547c02d90f481ddb2e8ea89dd2fbea498af990c5238ebdfe7"],"state_sha256":"39f7717c3029150f452c7f7b0b619652f098c14b55238e10d273079e5307f454"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R3JJcUbv7Qw1zkFVFAypoLSjzdmTe+DhW6czwF2ixoP1K/9VF69I7tnEBlZbAUbLgvpTO1QVdSJWrl0R2I5wCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T17:13:06.207740Z","bundle_sha256":"45c0e8b337ccbdc10c8a753116cd7a996eea3efd343a499a297fc63a91dcd5ec"}}