{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:H5VPMRH3ZWEZIRVA3JURJ36NSU","short_pith_number":"pith:H5VPMRH3","schema_version":"1.0","canonical_sha256":"3f6af644fbcd899446a0da6914efcd951095b5d33afb0a260046634726373acb","source":{"kind":"arxiv","id":"2605.15198","version":1},"attestation_state":"computed","paper":{"title":"ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"One discrete functional token suffices for both agentic operations and latent visual reasoning.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Pheng-Ann Heng, Rain Liu, Xinyan Chen, Ziyu Guo","submitted_at":"2026-05-14T17:59:55Z","abstract_excerpt":"Visual reasoning, often interleaved with intermediate visual states, has emerged as a promising direction in the field. A straightforward approach is to directly generate images via unified models during reasoning, but this is computationally expensive and architecturally non-trivial. Recent alternatives include agentic reasoning through code or tool calls, and latent reasoning with learnable hidden embeddings. However, agentic methods incur context-switching latency from external execution, while latent methods lack task generalization and are difficult to train with autoregressive paralleliz"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.15198","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-14T17:59:55Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"402a185441e65509516ec4e2243c4b86cb4d33c0e299ea49cffc3afaf9d513b1","abstract_canon_sha256":"c1c2ced4f400f5c1fce21efc76027fb18be7750b51ced415b4e1c73bacb6dbbc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.2","canonical_sha256":"3f6af644fbcd899446a0da6914efcd951095b5d33afb0a260046634726373acb","last_reissued_at":"2026-05-17T21:57:18.377459Z","signature_status":"unsigned_v0","first_computed_at":"2026-05-17T21:40:24.984558Z"},"graph_snapshot":{"paper":{"title":"ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"One discrete functional token suffices for both agentic operations and latent visual reasoning.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Pheng-Ann Heng, Rain Liu, Xinyan Chen, Ziyu Guo","submitted_at":"2026-05-14T17:59:55Z","abstract_excerpt":"Visual reasoning, often interleaved with intermediate visual states, has emerged as a promising direction in the field. A straightforward approach is to directly generate images via unified models during reasoning, but this is computationally expensive and architecturally non-trivial. Recent alternatives include agentic reasoning through code or tool calls, and latent reasoning with learnable hidden embeddings. However, agentic methods incur context-switching latency from external execution, while latent methods lack task generalization and are difficult to train with autoregressive paralleliz"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ATLAS achieves superior performance on challenging benchmarks while maintaining clear interpretability by using a single discrete functional token that serves as both an agentic operation and a latent visual reasoning unit.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a single discrete functional token can effectively internalize visual operations without any visual supervision and still generalize across tasks when generated via standard next-token prediction.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ATLAS uses a single functional token to unify agentic and latent visual reasoning without image generation or external execution.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"One discrete functional token suffices for both agentic operations and latent visual reasoning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"54db4534ab32f0ea5bf9dc95c6c87a3c4b3e69a860ce4d6570c579b438f4a032"},"source":{"id":"2605.15198","kind":"arxiv","version":1},"verdict":{"id":"9f14be86-a865-4e9e-9c3a-131216c39a5d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T03:05:18.872222Z","strongest_claim":"ATLAS achieves superior performance on challenging benchmarks while maintaining clear interpretability by using a single discrete functional token that serves as both an agentic operation and a latent visual reasoning unit.","one_line_summary":"ATLAS uses a single functional token to unify agentic and latent visual reasoning without image generation or external execution.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a single discrete functional token can effectively internalize visual operations without any visual supervision and still generalize across tasks when generated via standard next-token prediction.","pith_extraction_headline":"One discrete functional token suffices for both agentic operations and latent visual reasoning."},"references":{"count":24,"sample":[{"doi":"","year":null,"title":"LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Training","work_id":"41c2802e-aff9-482f-b506-10955ff0838d","ref_index":1,"cited_arxiv_id":"2509.23661","is_internal_anchor":true},{"doi":"","year":null,"title":"Qwen3-VL Technical Report","work_id":"1fe243aa-e3c0-4da6-b391-4cbcfc88d5c0","ref_index":2,"cited_arxiv_id":"2511.21631","is_internal_anchor":true},{"doi":"","year":null,"title":"Anole: An open, autoregressive, native large multimodal models for interleaved image-text generation","work_id":"31ce9d99-2071-41a0-9f51-51b8c5e3ba7e","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities","work_id":"008df105-2fdd-45d8-857a-8e35868aecb6","ref_index":4,"cited_arxiv_id":"2507.06261","is_internal_anchor":true},{"doi":"","year":null,"title":"Emerging Properties in Unified Multimodal Pretraining","work_id":"e0cfd82c-f5d4-44fd-b531-ec73ab0a805b","ref_index":5,"cited_arxiv_id":"2505.14683","is_internal_anchor":true}],"resolved_work":24,"snapshot_sha256":"59531425264020a4256c48ab840b47299f63cfc2a769c710677630731a0adee2","internal_anchors":18},"formal_canon":{"evidence_count":2,"snapshot_sha256":"affdd43ea14ce12d209cda4b098642c2992f038d3c83e22bfc1f7f6cec067d5e"},"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":"2605.15198","created_at":"2026-05-17T21:18:32.693370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.15198v1","created_at":"2026-05-17T21:18:32.693370+00:00"},{"alias_kind":"pith_short_12","alias_value":"H5VPMRH3ZWEZ","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"H5VPMRH3ZWEZIRVA","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"H5VPMRH3","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08763","citing_title":"OpenCoF: Learning to Reason Through Video Generation","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU","json":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU.json","graph_json":"https://pith.science/api/pith-number/H5VPMRH3ZWEZIRVA3JURJ36NSU/graph.json","events_json":"https://pith.science/api/pith-number/H5VPMRH3ZWEZIRVA3JURJ36NSU/events.json","paper":"https://pith.science/paper/H5VPMRH3"},"agent_actions":{"view_html":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU","download_json":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU.json","view_paper":"https://pith.science/paper/H5VPMRH3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.15198&json=true","fetch_graph":"https://pith.science/api/pith-number/H5VPMRH3ZWEZIRVA3JURJ36NSU/graph.json","fetch_events":"https://pith.science/api/pith-number/H5VPMRH3ZWEZIRVA3JURJ36NSU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU/action/storage_attestation","attest_author":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU/action/author_attestation","sign_citation":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU/action/citation_signature","submit_replication":"https://pith.science/pith/H5VPMRH3ZWEZIRVA3JURJ36NSU/action/replication_record"}},"created_at":"2026-05-17T21:18:32.693370+00:00","updated_at":"2026-05-17T21:57:18.377583+00:00"}