{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DFX6G257KKH7VJYXQBFG7BL5UL","short_pith_number":"pith:DFX6G257","schema_version":"1.0","canonical_sha256":"196fe36bbf528ffaa717804a6f857da2de395b2d569d005e83b88d58bab01d82","source":{"kind":"arxiv","id":"2511.20644","version":2},"attestation_state":"computed","paper":{"title":"Vision-Language Memory for Spatial Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Wang, Cherie Ho, Shaoshu Su, Taimeng Fu, Yi Du, Zuntao Liu","submitted_at":"2025-11-25T18:59:02Z","abstract_excerpt":"Spatial reasoning is a critical capability for intelligent robots, yet current vision-language models (VLMs) still fall short of human-level performance in video-based spatial reasoning. This gap mainly stems from two challenges: a semantic-geometric misalignment that prevents consistent 3D understanding, and the absence of persistent memory to retain 3D representation and understanding across frames. To address these limitations, we present VLM$^2$, a Vision-Language Model with persistent Memory for spatial reasoning with a view-consistent, 3D-aware representation purely from 2D videos. Speci"},"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":"2511.20644","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-11-25T18:59:02Z","cross_cats_sorted":[],"title_canon_sha256":"e0c6d2a18e9a74b9485e5442fae926cfa972f990e8e6f3ecbbbbac84722a1a9f","abstract_canon_sha256":"1b47657cd2e11e59c13340683030738cf3f09cc2fded3088b90abca3a209aa41"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-10T01:19:40.855204Z","signature_b64":"Qql8P4BiTzm+zp0rugfSveqGiCsWr0W7s1YKNC6nSmp+0isaby/7OZ1X0ITt/GLguTsYoPkDFqagsCprASU4BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"196fe36bbf528ffaa717804a6f857da2de395b2d569d005e83b88d58bab01d82","last_reissued_at":"2026-07-10T01:19:40.854650Z","signature_status":"signed_v1","first_computed_at":"2026-07-10T01:19:40.854650Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vision-Language Memory for Spatial Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Wang, Cherie Ho, Shaoshu Su, Taimeng Fu, Yi Du, Zuntao Liu","submitted_at":"2025-11-25T18:59:02Z","abstract_excerpt":"Spatial reasoning is a critical capability for intelligent robots, yet current vision-language models (VLMs) still fall short of human-level performance in video-based spatial reasoning. This gap mainly stems from two challenges: a semantic-geometric misalignment that prevents consistent 3D understanding, and the absence of persistent memory to retain 3D representation and understanding across frames. To address these limitations, we present VLM$^2$, a Vision-Language Model with persistent Memory for spatial reasoning with a view-consistent, 3D-aware representation purely from 2D videos. Speci"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.20644","kind":"arxiv","version":2},"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/2511.20644/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":"2511.20644","created_at":"2026-07-10T01:19:40.854717+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.20644v2","created_at":"2026-07-10T01:19:40.854717+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.20644","created_at":"2026-07-10T01:19:40.854717+00:00"},{"alias_kind":"pith_short_12","alias_value":"DFX6G257KKH7","created_at":"2026-07-10T01:19:40.854717+00:00"},{"alias_kind":"pith_short_16","alias_value":"DFX6G257KKH7VJYX","created_at":"2026-07-10T01:19:40.854717+00:00"},{"alias_kind":"pith_short_8","alias_value":"DFX6G257","created_at":"2026-07-10T01:19:40.854717+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2606.20905","citing_title":"Vesta: A Generalist Embodied Reasoning Model","ref_index":75,"is_internal_anchor":true},{"citing_arxiv_id":"2605.27318","citing_title":"Q-GeoMem: Question-Guided Geometric Memory for Video Spatial Reasoning","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2605.31196","citing_title":"Probing Collision Grounding in Vision-Language Models for Safe Human-Robot Collaboration","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18209","citing_title":"SPATIOROUTE: Dynamic Prompt Routing for Zero-Shot Spatial Reasoning","ref_index":17,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL","json":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL.json","graph_json":"https://pith.science/api/pith-number/DFX6G257KKH7VJYXQBFG7BL5UL/graph.json","events_json":"https://pith.science/api/pith-number/DFX6G257KKH7VJYXQBFG7BL5UL/events.json","paper":"https://pith.science/paper/DFX6G257"},"agent_actions":{"view_html":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL","download_json":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL.json","view_paper":"https://pith.science/paper/DFX6G257","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.20644&json=true","fetch_graph":"https://pith.science/api/pith-number/DFX6G257KKH7VJYXQBFG7BL5UL/graph.json","fetch_events":"https://pith.science/api/pith-number/DFX6G257KKH7VJYXQBFG7BL5UL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL/action/storage_attestation","attest_author":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL/action/author_attestation","sign_citation":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL/action/citation_signature","submit_replication":"https://pith.science/pith/DFX6G257KKH7VJYXQBFG7BL5UL/action/replication_record"}},"created_at":"2026-07-10T01:19:40.854717+00:00","updated_at":"2026-07-10T01:19:40.854717+00:00"}