{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3KLKFESYVK37GYYTFIS3M37HVQ","short_pith_number":"pith:3KLKFESY","schema_version":"1.0","canonical_sha256":"da96a29258aab7f363132a25b66fe7ac33a251c6a0390e1679ab3f619ec87a7e","source":{"kind":"arxiv","id":"2605.30904","version":1},"attestation_state":"computed","paper":{"title":"MergeTok: Unified Continuous and Discrete Visual Tokenization via Token Merging","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anna Wang, Chen Chen, Cheng Tan, Haonan Lu, Haoqian Wang, Luyuan Zhang, Qingsong Xie, Siyuan Li, Yanhao Zhang, Zedong Wang","submitted_at":"2026-05-29T06:38:10Z","abstract_excerpt":"Most visual tokenizers for image generation are bifurcated into two families with complementary limitations: continuous VAEs offer high-fidelity reconstruction but suffer from dense, entangled latents that are poorly suited for semantic control, whereas discrete VQ-based models enable autoregressive generation yet struggle with gradient sparsity, unstable training, and codebook collapse. In this work, we introduce MergeTok, a unified tokenizer that jointly optimizes continuous (VAE) and discrete (VQ) tokenizers within a encoder-decoder architecture, leveraging token merging techniques as a sem"},"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":"2605.30904","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-29T06:38:10Z","cross_cats_sorted":[],"title_canon_sha256":"155086a2953f9cf2ba319f2530b079cae4ceb28809c83abfd218a351e824c198","abstract_canon_sha256":"56f9ef68a9d87d3d685570936638e09327767fb2b3deb806dbe5bb3933d9b5d4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:24.845272Z","signature_b64":"wfkpvbvK8cq2iuAa5wobUN/k4uaOYEOJyFWrzisCU5tuOXQkDEoYw9YB7jpuMZ/O/H/aSQccGEWkOk5IccpICA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da96a29258aab7f363132a25b66fe7ac33a251c6a0390e1679ab3f619ec87a7e","last_reissued_at":"2026-06-01T01:03:24.844258Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:24.844258Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MergeTok: Unified Continuous and Discrete Visual Tokenization via Token Merging","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anna Wang, Chen Chen, Cheng Tan, Haonan Lu, Haoqian Wang, Luyuan Zhang, Qingsong Xie, Siyuan Li, Yanhao Zhang, Zedong Wang","submitted_at":"2026-05-29T06:38:10Z","abstract_excerpt":"Most visual tokenizers for image generation are bifurcated into two families with complementary limitations: continuous VAEs offer high-fidelity reconstruction but suffer from dense, entangled latents that are poorly suited for semantic control, whereas discrete VQ-based models enable autoregressive generation yet struggle with gradient sparsity, unstable training, and codebook collapse. In this work, we introduce MergeTok, a unified tokenizer that jointly optimizes continuous (VAE) and discrete (VQ) tokenizers within a encoder-decoder architecture, leveraging token merging techniques as a sem"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30904","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/2605.30904/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":"2605.30904","created_at":"2026-06-01T01:03:24.844431+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30904v1","created_at":"2026-06-01T01:03:24.844431+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30904","created_at":"2026-06-01T01:03:24.844431+00:00"},{"alias_kind":"pith_short_12","alias_value":"3KLKFESYVK37","created_at":"2026-06-01T01:03:24.844431+00:00"},{"alias_kind":"pith_short_16","alias_value":"3KLKFESYVK37GYYT","created_at":"2026-06-01T01:03:24.844431+00:00"},{"alias_kind":"pith_short_8","alias_value":"3KLKFESY","created_at":"2026-06-01T01:03:24.844431+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/3KLKFESYVK37GYYTFIS3M37HVQ","json":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ.json","graph_json":"https://pith.science/api/pith-number/3KLKFESYVK37GYYTFIS3M37HVQ/graph.json","events_json":"https://pith.science/api/pith-number/3KLKFESYVK37GYYTFIS3M37HVQ/events.json","paper":"https://pith.science/paper/3KLKFESY"},"agent_actions":{"view_html":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ","download_json":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ.json","view_paper":"https://pith.science/paper/3KLKFESY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30904&json=true","fetch_graph":"https://pith.science/api/pith-number/3KLKFESYVK37GYYTFIS3M37HVQ/graph.json","fetch_events":"https://pith.science/api/pith-number/3KLKFESYVK37GYYTFIS3M37HVQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ/action/storage_attestation","attest_author":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ/action/author_attestation","sign_citation":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ/action/citation_signature","submit_replication":"https://pith.science/pith/3KLKFESYVK37GYYTFIS3M37HVQ/action/replication_record"}},"created_at":"2026-06-01T01:03:24.844431+00:00","updated_at":"2026-06-01T01:03:24.844431+00:00"}