{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2EBPGS53TJZ45COJTUGC6FT4XN","short_pith_number":"pith:2EBPGS53","schema_version":"1.0","canonical_sha256":"d102f34bbb9a73ce89c99d0c2f167cbb6239a572b11b76106c421562dacd372c","source":{"kind":"arxiv","id":"2505.16839","version":3},"attestation_state":"computed","paper":{"title":"LaViDa: A Large Diffusion Language Model for Multimodal Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Aditya Grover, Akash Gokul, Hritik Bansal, Jason Kuen, Kai-Wei Chang, Kazuki Kozuka, Konstantinos Kallidromitis, Shufan Li, Yusuke Kato, Zhe Lin","submitted_at":"2025-05-22T16:07:12Z","abstract_excerpt":"Modern Vision-Language Models (VLMs) can solve a wide range of tasks requiring visual reasoning. In real-world scenarios, desirable properties for VLMs include fast inference and controllable generation (e.g., constraining outputs to adhere to a desired format). However, existing autoregressive (AR) VLMs like LLaVA struggle in these aspects. Discrete diffusion models (DMs) offer a promising alternative, enabling parallel decoding for faster inference and bidirectional context for controllable generation through text-infilling. While effective in language-only settings, DMs' potential for multi"},"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":"2505.16839","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-05-22T16:07:12Z","cross_cats_sorted":[],"title_canon_sha256":"a945a022313c5c3496bac58405254991bd8c680aa55642b8450dee2023e38c52","abstract_canon_sha256":"ccf97336654664a4bc2719a43d26f34833aabeddffa7b56978e6f2146ea8618d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:28.458458Z","signature_b64":"e2yHZE28tABHatTusBMXcCDpQtKE1URpGaFPPqYtPSacRxNCDON+CYU2Dj+BZ7HihKj2PWU59GYunCfbR3GVDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d102f34bbb9a73ce89c99d0c2f167cbb6239a572b11b76106c421562dacd372c","last_reissued_at":"2026-07-05T11:23:28.457929Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:28.457929Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LaViDa: A Large Diffusion Language Model for Multimodal Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Aditya Grover, Akash Gokul, Hritik Bansal, Jason Kuen, Kai-Wei Chang, Kazuki Kozuka, Konstantinos Kallidromitis, Shufan Li, Yusuke Kato, Zhe Lin","submitted_at":"2025-05-22T16:07:12Z","abstract_excerpt":"Modern Vision-Language Models (VLMs) can solve a wide range of tasks requiring visual reasoning. In real-world scenarios, desirable properties for VLMs include fast inference and controllable generation (e.g., constraining outputs to adhere to a desired format). However, existing autoregressive (AR) VLMs like LLaVA struggle in these aspects. Discrete diffusion models (DMs) offer a promising alternative, enabling parallel decoding for faster inference and bidirectional context for controllable generation through text-infilling. While effective in language-only settings, DMs' potential for multi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.16839","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/2505.16839/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":"2505.16839","created_at":"2026-07-05T11:23:28.457988+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.16839v3","created_at":"2026-07-05T11:23:28.457988+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.16839","created_at":"2026-07-05T11:23:28.457988+00:00"},{"alias_kind":"pith_short_12","alias_value":"2EBPGS53TJZ4","created_at":"2026-07-05T11:23:28.457988+00:00"},{"alias_kind":"pith_short_16","alias_value":"2EBPGS53TJZ45COJ","created_at":"2026-07-05T11:23:28.457988+00:00"},{"alias_kind":"pith_short_8","alias_value":"2EBPGS53","created_at":"2026-07-05T11:23:28.457988+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05722","citing_title":"Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decoding","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2606.29814","citing_title":"Nemotron-Labs-Diffusion-Image: Advancing Masked Discrete Diffusion for High-Resolution Image Synthesis","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30876","citing_title":"dMoE: dLLMs with Learnable Block Experts","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2512.21815","citing_title":"High-Entropy Tokens as Multimodal Failure Points in Vision-Language Models","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2512.15745","citing_title":"LLaDA2.0: Scaling Up Diffusion Language Models to 100B","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09302","citing_title":"Discrete Langevin-Inspired Posterior Sampling","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07399","citing_title":"GPO-V: Jailbreak Diffusion Vision Language Model by Global Probability Optimization","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22152","citing_title":"dWorldEval: Scalable Robotic Policy Evaluation via Discrete Diffusion World Model","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07399","citing_title":"GPO-V: Jailbreak Diffusion Vision Language Model by Global Probability Optimization","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05497","citing_title":"Thinking Diffusion: Penalize and Guide Visual-Grounded Reasoning in Diffusion Multimodal Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16514","citing_title":"BARD: Bridging AutoRegressive and Diffusion Vision-Language Models Via Highly Efficient Progressive Block Merging and Stage-Wise Distillation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20796","citing_title":"LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN","json":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN.json","graph_json":"https://pith.science/api/pith-number/2EBPGS53TJZ45COJTUGC6FT4XN/graph.json","events_json":"https://pith.science/api/pith-number/2EBPGS53TJZ45COJTUGC6FT4XN/events.json","paper":"https://pith.science/paper/2EBPGS53"},"agent_actions":{"view_html":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN","download_json":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN.json","view_paper":"https://pith.science/paper/2EBPGS53","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.16839&json=true","fetch_graph":"https://pith.science/api/pith-number/2EBPGS53TJZ45COJTUGC6FT4XN/graph.json","fetch_events":"https://pith.science/api/pith-number/2EBPGS53TJZ45COJTUGC6FT4XN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN/action/storage_attestation","attest_author":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN/action/author_attestation","sign_citation":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN/action/citation_signature","submit_replication":"https://pith.science/pith/2EBPGS53TJZ45COJTUGC6FT4XN/action/replication_record"}},"created_at":"2026-07-05T11:23:28.457988+00:00","updated_at":"2026-07-05T11:23:28.457988+00:00"}