{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XYNGT7GRJZMQQ7WMPE3AGKI6FF","short_pith_number":"pith:XYNGT7GR","schema_version":"1.0","canonical_sha256":"be1a69fcd14e59087ecc793603291e2969c958cc1ca8b909d03039c69c366880","source":{"kind":"arxiv","id":"2509.01400","version":1},"attestation_state":"computed","paper":{"title":"Distillation of a tractable model from the VQ-VAE","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Armin Had\\v{z}i\\'c, Milan Papez, Tom\\'a\\v{s} Pevn\\'y","submitted_at":"2025-09-01T11:51:08Z","abstract_excerpt":"Deep generative models with discrete latent space, such as the Vector-Quantized Variational Autoencoder (VQ-VAE), offer excellent data generation capabilities, but, due to the large size of their latent space, their probabilistic inference is deemed intractable. We demonstrate that the VQ-VAE can be distilled into a tractable model by selecting a subset of latent variables with high probabilities. This simple strategy is particularly efficient, especially if the VQ-VAE underutilizes its latent space, which is, indeed, very often the case. We frame the distilled model as a probabilistic circuit"},"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":"2509.01400","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-09-01T11:51:08Z","cross_cats_sorted":[],"title_canon_sha256":"ba8b51ce7d3e26472aa477210362f54f357928be9a504a7af44a6eda705b6a37","abstract_canon_sha256":"c2c46bbcfd7f6dfef80dc6b67c9578cab457da9608a3de96dc4b56e9ade79bc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:02:59.868972Z","signature_b64":"raqgAypfKA+szSn1oZBlXlTQEIU4iclHnM7t6l3HNCMctHCMJ4W1zXRruT/D1rqUpIlxoCUcVkk9GGOVJwDpBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"be1a69fcd14e59087ecc793603291e2969c958cc1ca8b909d03039c69c366880","last_reissued_at":"2026-07-05T12:02:59.868482Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:02:59.868482Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distillation of a tractable model from the VQ-VAE","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Armin Had\\v{z}i\\'c, Milan Papez, Tom\\'a\\v{s} Pevn\\'y","submitted_at":"2025-09-01T11:51:08Z","abstract_excerpt":"Deep generative models with discrete latent space, such as the Vector-Quantized Variational Autoencoder (VQ-VAE), offer excellent data generation capabilities, but, due to the large size of their latent space, their probabilistic inference is deemed intractable. We demonstrate that the VQ-VAE can be distilled into a tractable model by selecting a subset of latent variables with high probabilities. This simple strategy is particularly efficient, especially if the VQ-VAE underutilizes its latent space, which is, indeed, very often the case. We frame the distilled model as a probabilistic circuit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.01400","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/2509.01400/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":"2509.01400","created_at":"2026-07-05T12:02:59.868542+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.01400v1","created_at":"2026-07-05T12:02:59.868542+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.01400","created_at":"2026-07-05T12:02:59.868542+00:00"},{"alias_kind":"pith_short_12","alias_value":"XYNGT7GRJZMQ","created_at":"2026-07-05T12:02:59.868542+00:00"},{"alias_kind":"pith_short_16","alias_value":"XYNGT7GRJZMQQ7WM","created_at":"2026-07-05T12:02:59.868542+00:00"},{"alias_kind":"pith_short_8","alias_value":"XYNGT7GR","created_at":"2026-07-05T12:02:59.868542+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/XYNGT7GRJZMQQ7WMPE3AGKI6FF","json":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF.json","graph_json":"https://pith.science/api/pith-number/XYNGT7GRJZMQQ7WMPE3AGKI6FF/graph.json","events_json":"https://pith.science/api/pith-number/XYNGT7GRJZMQQ7WMPE3AGKI6FF/events.json","paper":"https://pith.science/paper/XYNGT7GR"},"agent_actions":{"view_html":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF","download_json":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF.json","view_paper":"https://pith.science/paper/XYNGT7GR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.01400&json=true","fetch_graph":"https://pith.science/api/pith-number/XYNGT7GRJZMQQ7WMPE3AGKI6FF/graph.json","fetch_events":"https://pith.science/api/pith-number/XYNGT7GRJZMQQ7WMPE3AGKI6FF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF/action/storage_attestation","attest_author":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF/action/author_attestation","sign_citation":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF/action/citation_signature","submit_replication":"https://pith.science/pith/XYNGT7GRJZMQQ7WMPE3AGKI6FF/action/replication_record"}},"created_at":"2026-07-05T12:02:59.868542+00:00","updated_at":"2026-07-05T12:02:59.868542+00:00"}