{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7OM5ZJNWDRB4IBZLSGXFMZDSNL","short_pith_number":"pith:7OM5ZJNW","schema_version":"1.0","canonical_sha256":"fb99dca5b61c43c4072b91ae5664726ad83b6b74bc5b0b708ec7c88c27ea917a","source":{"kind":"arxiv","id":"2606.27147","version":1},"attestation_state":"computed","paper":{"title":"Safe Autoregressive Image Generation with Iterative Self-Improving Codebooks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jindong Gu, Philip Torr, Yunqi Xue, Zhijiang Li","submitted_at":"2026-06-25T15:18:31Z","abstract_excerpt":"Unlike diffusion-based models that operate in continuous latent spaces, autoregressive unified multimodal models produce images by sequentially predicting discretized visual tokens. These tokens are derived from a codebook that maps embeddings to quantized visual patterns. The language-like architecture enables unified multimodal models to effectively capture text conditional information for generation, making them promising for text-to-image tasks. This also raises an interesting question: how safe are the images generated in such an autoregressive way? In this work, we propose iterative self"},"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":"2606.27147","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-06-25T15:18:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d5328fdfdaf6d0d548a8d1bbfbb3e5a847589fd645e4d6280cf27155c3bd968d","abstract_canon_sha256":"70d1d4712191f3d3aae1d3d2fcf8d6ce06677abf785b785f205c10bf4e83c56f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:16:11.520241Z","signature_b64":"04RhWVcSe6xU5GdRZwvGl3E8jDtOu0EglDce0DEUWEzC1mseNTX4IZsLZQm0WLwCM9+NeD8M3hF2Skw4V4SdDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fb99dca5b61c43c4072b91ae5664726ad83b6b74bc5b0b708ec7c88c27ea917a","last_reissued_at":"2026-06-26T01:16:11.519860Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:16:11.519860Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safe Autoregressive Image Generation with Iterative Self-Improving Codebooks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Jindong Gu, Philip Torr, Yunqi Xue, Zhijiang Li","submitted_at":"2026-06-25T15:18:31Z","abstract_excerpt":"Unlike diffusion-based models that operate in continuous latent spaces, autoregressive unified multimodal models produce images by sequentially predicting discretized visual tokens. These tokens are derived from a codebook that maps embeddings to quantized visual patterns. The language-like architecture enables unified multimodal models to effectively capture text conditional information for generation, making them promising for text-to-image tasks. This also raises an interesting question: how safe are the images generated in such an autoregressive way? In this work, we propose iterative self"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27147","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/2606.27147/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":"2606.27147","created_at":"2026-06-26T01:16:11.519914+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.27147v1","created_at":"2026-06-26T01:16:11.519914+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27147","created_at":"2026-06-26T01:16:11.519914+00:00"},{"alias_kind":"pith_short_12","alias_value":"7OM5ZJNWDRB4","created_at":"2026-06-26T01:16:11.519914+00:00"},{"alias_kind":"pith_short_16","alias_value":"7OM5ZJNWDRB4IBZL","created_at":"2026-06-26T01:16:11.519914+00:00"},{"alias_kind":"pith_short_8","alias_value":"7OM5ZJNW","created_at":"2026-06-26T01:16:11.519914+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/7OM5ZJNWDRB4IBZLSGXFMZDSNL","json":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL.json","graph_json":"https://pith.science/api/pith-number/7OM5ZJNWDRB4IBZLSGXFMZDSNL/graph.json","events_json":"https://pith.science/api/pith-number/7OM5ZJNWDRB4IBZLSGXFMZDSNL/events.json","paper":"https://pith.science/paper/7OM5ZJNW"},"agent_actions":{"view_html":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL","download_json":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL.json","view_paper":"https://pith.science/paper/7OM5ZJNW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.27147&json=true","fetch_graph":"https://pith.science/api/pith-number/7OM5ZJNWDRB4IBZLSGXFMZDSNL/graph.json","fetch_events":"https://pith.science/api/pith-number/7OM5ZJNWDRB4IBZLSGXFMZDSNL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL/action/storage_attestation","attest_author":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL/action/author_attestation","sign_citation":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL/action/citation_signature","submit_replication":"https://pith.science/pith/7OM5ZJNWDRB4IBZLSGXFMZDSNL/action/replication_record"}},"created_at":"2026-06-26T01:16:11.519914+00:00","updated_at":"2026-06-26T01:16:11.519914+00:00"}