{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:POWNTUBRHIUTSIXCQD6F6OHB3O","short_pith_number":"pith:POWNTUBR","schema_version":"1.0","canonical_sha256":"7bacd9d0313a293922e280fc5f38e1db949fd891a918fb7ff24fe3ef02dac93b","source":{"kind":"arxiv","id":"2505.07538","version":3},"attestation_state":"computed","paper":{"title":"Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bohan Wang, Fengda Zhang, Hanwang Zhang, Jiachun Pan, Junzhe Zhang, Kaihang Pan, Kennard Yanting Chan, Li'an Bi, Liyu Jia, Mingze Zhou, Saining Zhang, Shuo Chen, Wang Lin, Weijia Wu, Wei Zhao, Wentao Hu, Xue Song, Zhongqi Yue","submitted_at":"2025-05-12T13:19:08Z","abstract_excerpt":"We completely discard the conventional spatial prior in image representation and introduce a novel discrete visual tokenizer: Self-consistency Tokenizer (Selftok). At its design core, we compose an autoregressive (AR) prior -- mirroring the causal structure of language -- into visual tokens by using the reverse diffusion process of image generation. The AR property makes Selftok fundamentally distinct from traditional spatial tokens in the following two key ways: - Selftok offers an elegant and minimalist approach to unify diffusion and AR for vision-language models (VLMs): By representing ima"},"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":"2505.07538","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-12T13:19:08Z","cross_cats_sorted":[],"title_canon_sha256":"ec2e265ccb40d51de5d40bc00667a11a73ae045da827dbc456476fc7e04635d7","abstract_canon_sha256":"da7320e71ad5bcca60ce428cd93870a4a7b020442717f054ff7210c2016e47b3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:07.800109Z","signature_b64":"5k8PiQD1CxJwFhFBC1z+Knr5iY8teHKyYrv+1DiKzR8mRmUskDAVe+Igwsao6CGJ682us1SNrdx2kjPSVkH7Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7bacd9d0313a293922e280fc5f38e1db949fd891a918fb7ff24fe3ef02dac93b","last_reissued_at":"2026-07-05T11:10:07.799610Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:07.799610Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Selftok: Discrete Visual Tokens of Autoregression, by Diffusion, and for Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bohan Wang, Fengda Zhang, Hanwang Zhang, Jiachun Pan, Junzhe Zhang, Kaihang Pan, Kennard Yanting Chan, Li'an Bi, Liyu Jia, Mingze Zhou, Saining Zhang, Shuo Chen, Wang Lin, Weijia Wu, Wei Zhao, Wentao Hu, Xue Song, Zhongqi Yue","submitted_at":"2025-05-12T13:19:08Z","abstract_excerpt":"We completely discard the conventional spatial prior in image representation and introduce a novel discrete visual tokenizer: Self-consistency Tokenizer (Selftok). At its design core, we compose an autoregressive (AR) prior -- mirroring the causal structure of language -- into visual tokens by using the reverse diffusion process of image generation. The AR property makes Selftok fundamentally distinct from traditional spatial tokens in the following two key ways: - Selftok offers an elegant and minimalist approach to unify diffusion and AR for vision-language models (VLMs): By representing ima"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.07538","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/2505.07538/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":"2505.07538","created_at":"2026-07-05T11:10:07.799670+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.07538v3","created_at":"2026-07-05T11:10:07.799670+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.07538","created_at":"2026-07-05T11:10:07.799670+00:00"},{"alias_kind":"pith_short_12","alias_value":"POWNTUBRHIUT","created_at":"2026-07-05T11:10:07.799670+00:00"},{"alias_kind":"pith_short_16","alias_value":"POWNTUBRHIUTSIXC","created_at":"2026-07-05T11:10:07.799670+00:00"},{"alias_kind":"pith_short_8","alias_value":"POWNTUBR","created_at":"2026-07-05T11:10:07.799670+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.13289","citing_title":"HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06137","citing_title":"Autoregressive Visual Generation Needs a Prologue","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30248","citing_title":"Your Data Manifold is Secretly a Reward Model: Shell-LCC for Text-to-Video Generation","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06137","citing_title":"Autoregressive Visual Generation Needs a Prologue","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15453","citing_title":"(1D) Ordered Tokens Enable Efficient Test-Time Search","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O","json":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O.json","graph_json":"https://pith.science/api/pith-number/POWNTUBRHIUTSIXCQD6F6OHB3O/graph.json","events_json":"https://pith.science/api/pith-number/POWNTUBRHIUTSIXCQD6F6OHB3O/events.json","paper":"https://pith.science/paper/POWNTUBR"},"agent_actions":{"view_html":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O","download_json":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O.json","view_paper":"https://pith.science/paper/POWNTUBR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.07538&json=true","fetch_graph":"https://pith.science/api/pith-number/POWNTUBRHIUTSIXCQD6F6OHB3O/graph.json","fetch_events":"https://pith.science/api/pith-number/POWNTUBRHIUTSIXCQD6F6OHB3O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O/action/storage_attestation","attest_author":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O/action/author_attestation","sign_citation":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O/action/citation_signature","submit_replication":"https://pith.science/pith/POWNTUBRHIUTSIXCQD6F6OHB3O/action/replication_record"}},"created_at":"2026-07-05T11:10:07.799670+00:00","updated_at":"2026-07-05T11:10:07.799670+00:00"}