{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:NVQ64Y2TMSWAKKL6FOEHVVWZ4N","short_pith_number":"pith:NVQ64Y2T","schema_version":"1.0","canonical_sha256":"6d61ee635364ac05297e2b887ad6d9e35c7b8d5a5113cb87b082fd047ecfdfa5","source":{"kind":"arxiv","id":"2310.02423","version":2},"attestation_state":"computed","paper":{"title":"Delta-AI: Local objectives for amortized inference in sparse graphical models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chen Sun, Dinghuai Zhang, Dragos Secrieru, Guillaume Lajoie, Hae Beom Lee, Jean-Pierre Falet, Nikolay Malkin, Thomas Jiralerspong, Yoshua Bengio","submitted_at":"2023-10-03T20:37:03Z","abstract_excerpt":"We present a new algorithm for amortized inference in sparse probabilistic graphical models (PGMs), which we call $\\Delta$-amortized inference ($\\Delta$-AI). Our approach is based on the observation that when the sampling of variables in a PGM is seen as a sequence of actions taken by an agent, sparsity of the PGM enables local credit assignment in the agent's policy learning objective. This yields a local constraint that can be turned into a local loss in the style of generative flow networks (GFlowNets) that enables off-policy training but avoids the need to instantiate all the random variab"},"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":"2310.02423","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-03T20:37:03Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"d7310098e5333f6948e2fb95f01ae0dbab4c293fa94cd904bf21ec50255e334f","abstract_canon_sha256":"8ad1f0336061e911e89dfc1025fa250af10ec9a5061fa33bea216d7a89379581"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:55:45.194969Z","signature_b64":"CvjuUMKJkyrz2+ZCwHf/fxXnSwQ6eBtBHv3gd7Ry9SnmqqyDubM/gDcqjmRPIsOr92SVqs97Oa9+K4+mWGxkCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d61ee635364ac05297e2b887ad6d9e35c7b8d5a5113cb87b082fd047ecfdfa5","last_reissued_at":"2026-07-05T07:55:45.194455Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:55:45.194455Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Delta-AI: Local objectives for amortized inference in sparse graphical models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chen Sun, Dinghuai Zhang, Dragos Secrieru, Guillaume Lajoie, Hae Beom Lee, Jean-Pierre Falet, Nikolay Malkin, Thomas Jiralerspong, Yoshua Bengio","submitted_at":"2023-10-03T20:37:03Z","abstract_excerpt":"We present a new algorithm for amortized inference in sparse probabilistic graphical models (PGMs), which we call $\\Delta$-amortized inference ($\\Delta$-AI). Our approach is based on the observation that when the sampling of variables in a PGM is seen as a sequence of actions taken by an agent, sparsity of the PGM enables local credit assignment in the agent's policy learning objective. This yields a local constraint that can be turned into a local loss in the style of generative flow networks (GFlowNets) that enables off-policy training but avoids the need to instantiate all the random variab"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02423","kind":"arxiv","version":2},"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/2310.02423/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":"2310.02423","created_at":"2026-07-05T07:55:45.194518+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02423v2","created_at":"2026-07-05T07:55:45.194518+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02423","created_at":"2026-07-05T07:55:45.194518+00:00"},{"alias_kind":"pith_short_12","alias_value":"NVQ64Y2TMSWA","created_at":"2026-07-05T07:55:45.194518+00:00"},{"alias_kind":"pith_short_16","alias_value":"NVQ64Y2TMSWAKKL6","created_at":"2026-07-05T07:55:45.194518+00:00"},{"alias_kind":"pith_short_8","alias_value":"NVQ64Y2T","created_at":"2026-07-05T07:55:45.194518+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.08644","citing_title":"Path-dependent Discrete Amortized Inference","ref_index":111,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N","json":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N.json","graph_json":"https://pith.science/api/pith-number/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/graph.json","events_json":"https://pith.science/api/pith-number/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/events.json","paper":"https://pith.science/paper/NVQ64Y2T"},"agent_actions":{"view_html":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N","download_json":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N.json","view_paper":"https://pith.science/paper/NVQ64Y2T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02423&json=true","fetch_graph":"https://pith.science/api/pith-number/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/graph.json","fetch_events":"https://pith.science/api/pith-number/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/action/storage_attestation","attest_author":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/action/author_attestation","sign_citation":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/action/citation_signature","submit_replication":"https://pith.science/pith/NVQ64Y2TMSWAKKL6FOEHVVWZ4N/action/replication_record"}},"created_at":"2026-07-05T07:55:45.194518+00:00","updated_at":"2026-07-05T07:55:45.194518+00:00"}