{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:IGPGCFACJNPZLKPQR4MEIRPTRN","short_pith_number":"pith:IGPGCFAC","schema_version":"1.0","canonical_sha256":"419e6114024b5f95a9f08f184445f38b4525f074d8b50ef341d2a369b30d5da1","source":{"kind":"arxiv","id":"2410.15184","version":1},"attestation_state":"computed","paper":{"title":"Action abstractions for amortized sampling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Emmanuel Bengio, Joseph D Viviano, L\\'ena N\\'ehale Ezzine, Micha{\\l} Koziarski, Moksh Jain, Nikolay Malkin, Oussama Boussif, Rim Assouel, Yoshua Bengio","submitted_at":"2024-10-19T19:22:50Z","abstract_excerpt":"As trajectories sampled by policies used by reinforcement learning (RL) and generative flow networks (GFlowNets) grow longer, credit assignment and exploration become more challenging, and the long planning horizon hinders mode discovery and generalization. The challenge is particularly pronounced in entropy-seeking RL methods, such as generative flow networks, where the agent must learn to sample from a structured distribution and discover multiple high-reward states, each of which take many steps to reach. To tackle this challenge, we propose an approach to incorporate the discovery of actio"},"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":"2410.15184","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-19T19:22:50Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b3b3b039a22ae5500ee13592d78274eda3a66a1681ed285b53380afa3c815394","abstract_canon_sha256":"5b03043ab475b017c1ce55b2202ef79f9be4a70be4ae1f0906494dbe80bcb392"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:23:16.327150Z","signature_b64":"k6Cub/NGlH/DpBEXZdByam2XlMEII7Y+zky2m9hS8bBWMj0w4qM6a5Ag4CdYBTnHO6BV9agN6BIYlMH+R8nfBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"419e6114024b5f95a9f08f184445f38b4525f074d8b50ef341d2a369b30d5da1","last_reissued_at":"2026-07-05T09:23:16.326522Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:23:16.326522Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Action abstractions for amortized sampling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Emmanuel Bengio, Joseph D Viviano, L\\'ena N\\'ehale Ezzine, Micha{\\l} Koziarski, Moksh Jain, Nikolay Malkin, Oussama Boussif, Rim Assouel, Yoshua Bengio","submitted_at":"2024-10-19T19:22:50Z","abstract_excerpt":"As trajectories sampled by policies used by reinforcement learning (RL) and generative flow networks (GFlowNets) grow longer, credit assignment and exploration become more challenging, and the long planning horizon hinders mode discovery and generalization. The challenge is particularly pronounced in entropy-seeking RL methods, such as generative flow networks, where the agent must learn to sample from a structured distribution and discover multiple high-reward states, each of which take many steps to reach. To tackle this challenge, we propose an approach to incorporate the discovery of actio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.15184","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/2410.15184/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":"2410.15184","created_at":"2026-07-05T09:23:16.326582+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.15184v1","created_at":"2026-07-05T09:23:16.326582+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.15184","created_at":"2026-07-05T09:23:16.326582+00:00"},{"alias_kind":"pith_short_12","alias_value":"IGPGCFACJNPZ","created_at":"2026-07-05T09:23:16.326582+00:00"},{"alias_kind":"pith_short_16","alias_value":"IGPGCFACJNPZLKPQ","created_at":"2026-07-05T09:23:16.326582+00:00"},{"alias_kind":"pith_short_8","alias_value":"IGPGCFAC","created_at":"2026-07-05T09:23:16.326582+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/IGPGCFACJNPZLKPQR4MEIRPTRN","json":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN.json","graph_json":"https://pith.science/api/pith-number/IGPGCFACJNPZLKPQR4MEIRPTRN/graph.json","events_json":"https://pith.science/api/pith-number/IGPGCFACJNPZLKPQR4MEIRPTRN/events.json","paper":"https://pith.science/paper/IGPGCFAC"},"agent_actions":{"view_html":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN","download_json":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN.json","view_paper":"https://pith.science/paper/IGPGCFAC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.15184&json=true","fetch_graph":"https://pith.science/api/pith-number/IGPGCFACJNPZLKPQR4MEIRPTRN/graph.json","fetch_events":"https://pith.science/api/pith-number/IGPGCFACJNPZLKPQR4MEIRPTRN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN/action/storage_attestation","attest_author":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN/action/author_attestation","sign_citation":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN/action/citation_signature","submit_replication":"https://pith.science/pith/IGPGCFACJNPZLKPQR4MEIRPTRN/action/replication_record"}},"created_at":"2026-07-05T09:23:16.326582+00:00","updated_at":"2026-07-05T09:23:16.326582+00:00"}