{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:25NIDUZ5XYSI7XNUPVMANUIYGN","short_pith_number":"pith:25NIDUZ5","schema_version":"1.0","canonical_sha256":"d75a81d33dbe248fddb47d5806d1183379cd24f9f9b5ef654fd1d3ca52c27d9a","source":{"kind":"arxiv","id":"2512.06609","version":3},"attestation_state":"computed","paper":{"title":"Training-Free Vector Quantization via Gaussian VAEs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Jiajun He, Jie Tang, Jose Miguel Hernandez-Lobato, Tongda Xu, Wendi Zheng, Yan Wang, Ya-Qin Zhang","submitted_at":"2025-12-07T00:57:58Z","abstract_excerpt":"Vector-quantized variational autoencoders (VQ-VAEs) are discrete autoencoders that compress images into discrete tokens. However, they are difficult to train due to discretization. In this paper, we propose a simple yet effective technique dubbed Gaussian Quant (GQ), which first trains a Gaussian VAE under certain constraints and then converts it into a VQ-VAE without additional training. For conversion, GQ generates random Gaussian noise as a codebook and finds the closest noise vector to the posterior mean. Theoretically, we prove that when the logarithm of the codebook size exceeds the bits"},"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":"2512.06609","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-07T00:57:58Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"5117fab40830a279965eadce92a3074f92e16d46970bb0b4dcab902f3744f8dd","abstract_canon_sha256":"fac72dbe52a76dbfa20dc638ea441d5984f40892ef6d9b474a848ec875c9c187"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:04:53.326157Z","signature_b64":"nISyYCUSwuRcSB4P7O4GYUnBHt1UuUNdBZ+qPiwqF7cCpvuXE/7dKndplTPvJxSZnceUn2IeflHNBJ1tS0JmAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d75a81d33dbe248fddb47d5806d1183379cd24f9f9b5ef654fd1d3ca52c27d9a","last_reissued_at":"2026-05-27T01:04:53.325514Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:04:53.325514Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training-Free Vector Quantization via Gaussian VAEs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Jiajun He, Jie Tang, Jose Miguel Hernandez-Lobato, Tongda Xu, Wendi Zheng, Yan Wang, Ya-Qin Zhang","submitted_at":"2025-12-07T00:57:58Z","abstract_excerpt":"Vector-quantized variational autoencoders (VQ-VAEs) are discrete autoencoders that compress images into discrete tokens. However, they are difficult to train due to discretization. In this paper, we propose a simple yet effective technique dubbed Gaussian Quant (GQ), which first trains a Gaussian VAE under certain constraints and then converts it into a VQ-VAE without additional training. For conversion, GQ generates random Gaussian noise as a codebook and finds the closest noise vector to the posterior mean. Theoretically, we prove that when the logarithm of the codebook size exceeds the bits"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.06609","kind":"arxiv","version":3},"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/2512.06609/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":"2512.06609","created_at":"2026-05-27T01:04:53.325617+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.06609v3","created_at":"2026-05-27T01:04:53.325617+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.06609","created_at":"2026-05-27T01:04:53.325617+00:00"},{"alias_kind":"pith_short_12","alias_value":"25NIDUZ5XYSI","created_at":"2026-05-27T01:04:53.325617+00:00"},{"alias_kind":"pith_short_16","alias_value":"25NIDUZ5XYSI7XNU","created_at":"2026-05-27T01:04:53.325617+00:00"},{"alias_kind":"pith_short_8","alias_value":"25NIDUZ5","created_at":"2026-05-27T01:04:53.325617+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/25NIDUZ5XYSI7XNUPVMANUIYGN","json":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN.json","graph_json":"https://pith.science/api/pith-number/25NIDUZ5XYSI7XNUPVMANUIYGN/graph.json","events_json":"https://pith.science/api/pith-number/25NIDUZ5XYSI7XNUPVMANUIYGN/events.json","paper":"https://pith.science/paper/25NIDUZ5"},"agent_actions":{"view_html":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN","download_json":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN.json","view_paper":"https://pith.science/paper/25NIDUZ5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.06609&json=true","fetch_graph":"https://pith.science/api/pith-number/25NIDUZ5XYSI7XNUPVMANUIYGN/graph.json","fetch_events":"https://pith.science/api/pith-number/25NIDUZ5XYSI7XNUPVMANUIYGN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN/action/storage_attestation","attest_author":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN/action/author_attestation","sign_citation":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN/action/citation_signature","submit_replication":"https://pith.science/pith/25NIDUZ5XYSI7XNUPVMANUIYGN/action/replication_record"}},"created_at":"2026-05-27T01:04:53.325617+00:00","updated_at":"2026-05-27T01:04:53.325617+00:00"}