{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WSX4VCGEDWO5IKJJIMT5JSJKNN","short_pith_number":"pith:WSX4VCGE","schema_version":"1.0","canonical_sha256":"b4afca88c41d9dd429294327d4c92a6b59ac1fbea24e403379d007acb4efe80b","source":{"kind":"arxiv","id":"2504.15280","version":2},"attestation_state":"computed","paper":{"title":"Seeing from Another Perspective: Evaluating Multi-View Understanding in MLLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Chenyu Wang, Chun-Hsiao Yeh, Ruoyu Wang, Shengbang Tong, Shenghua Gao, Ta-Ying Cheng, Tianzhe Chu, Yi Ma, Yubei Chen, Yuexiang Zhai","submitted_at":"2025-04-21T17:59:53Z","abstract_excerpt":"Multi-view understanding, the ability to reconcile visual information across diverse viewpoints for effective navigation, manipulation, and 3D scene comprehension, is a fundamental challenge in Multi-Modal Large Language Models (MLLMs) to be used as embodied agents. While recent MLLMs have shown impressive advances in high-level reasoning and planning, they frequently fall short when confronted with multi-view geometric consistency and cross-view correspondence. To comprehensively evaluate the challenges of MLLMs in multi-view scene reasoning, we propose All-Angles Bench, a benchmark of over 2"},"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":"2504.15280","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-21T17:59:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"e24c4b352175d2a0b9d581fcbb26375af261ea2a66c78a956172d4fd568dbd74","abstract_canon_sha256":"e319e6d81d3a1760bc8ac6e2bea57e8e58343818a659fcb296748580e0d8400c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:54:47.277959Z","signature_b64":"+OGKMx5H5tPgYJR04E+G3j9rRIf0L/FnLPJzBoh6SDnkN6X00DjKJw7h6mXijO84vnXIv0zpaTCH3kp5DFZPCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b4afca88c41d9dd429294327d4c92a6b59ac1fbea24e403379d007acb4efe80b","last_reissued_at":"2026-07-05T10:54:47.277444Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:54:47.277444Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Seeing from Another Perspective: Evaluating Multi-View Understanding in MLLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Chenyu Wang, Chun-Hsiao Yeh, Ruoyu Wang, Shengbang Tong, Shenghua Gao, Ta-Ying Cheng, Tianzhe Chu, Yi Ma, Yubei Chen, Yuexiang Zhai","submitted_at":"2025-04-21T17:59:53Z","abstract_excerpt":"Multi-view understanding, the ability to reconcile visual information across diverse viewpoints for effective navigation, manipulation, and 3D scene comprehension, is a fundamental challenge in Multi-Modal Large Language Models (MLLMs) to be used as embodied agents. While recent MLLMs have shown impressive advances in high-level reasoning and planning, they frequently fall short when confronted with multi-view geometric consistency and cross-view correspondence. To comprehensively evaluate the challenges of MLLMs in multi-view scene reasoning, we propose All-Angles Bench, a benchmark of over 2"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.15280","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/2504.15280/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":"2504.15280","created_at":"2026-07-05T10:54:47.277506+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.15280v2","created_at":"2026-07-05T10:54:47.277506+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.15280","created_at":"2026-07-05T10:54:47.277506+00:00"},{"alias_kind":"pith_short_12","alias_value":"WSX4VCGEDWO5","created_at":"2026-07-05T10:54:47.277506+00:00"},{"alias_kind":"pith_short_16","alias_value":"WSX4VCGEDWO5IKJJ","created_at":"2026-07-05T10:54:47.277506+00:00"},{"alias_kind":"pith_short_8","alias_value":"WSX4VCGE","created_at":"2026-07-05T10:54:47.277506+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":27,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26029","citing_title":"TriViewBench: Controlled Complexity Scaling for Multi-View Structural Reasoning in MLLMs","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25634","citing_title":"SSMNBench: Diagnosing Image-based Cross-View Human-Object Understanding via Single-View Sufficiency and Multi-View Necessity","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22694","citing_title":"SATURN: Symbolic Spatial Reasoning for Multi-Perspective Grounding","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20905","citing_title":"Vesta: A Generalist Embodied Reasoning Model","ref_index":144,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10651","citing_title":"Kwai Keye-VL-2.0 Technical Report","ref_index":102,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00248","citing_title":"Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity","ref_index":130,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05677","citing_title":"LongSpace: Exploring Long-Horizon Spatial Memory from Perception to Recall in Video","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03988","citing_title":"Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03577","citing_title":"Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02742","citing_title":"Consistent Yet Wrong: Evidence Insensitivity in Spatial Vision-Language Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02482","citing_title":"X-Stream: Exploring MLLMs as Multiplexers for Multi-Stream Understanding","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23176","citing_title":"DRIVESPATIAL: A Benchmark for Spatiotemporal Intelligence in VLMs for Autonomous Driving","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02482","citing_title":"X-Stream: Exploring MLLMs as Multiplexers for Multi-Stream Understanding","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30231","citing_title":"Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23176","citing_title":"DRIVESPATIAL: A Benchmark for Spatiotemporal Intelligence in VLMs for Autonomous Driving","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22819","citing_title":"Cambrian-P: Pose-Grounded Video Understanding","ref_index":115,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04128","citing_title":"JoyAI-Image: Awaking Spatial Intelligence in Unified Multimodal Understanding and Generation","ref_index":98,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18621","citing_title":"CrossView Suite: Harnessing Cross-view Spatial Intelligence of MLLMs with Dataset, Model and Benchmark","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2505.21649","citing_title":"Seeing Isn't Orienting: A Cognitively Grounded Benchmark Reveals Systematic Orientation Failures in MLLMs","ref_index":121,"is_internal_anchor":false},{"citing_arxiv_id":"2511.04670","citing_title":"Cambrian-S: Towards Spatial Supersensing in Video","ref_index":152,"is_internal_anchor":false},{"citing_arxiv_id":"2512.23365","citing_title":"SpatialMosaic: A Multiview VLM Dataset for Partial Visibility","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2603.03944","citing_title":"SCP: Spatial Causal Prediction in Video","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02130","citing_title":"From Where Things Are to What They Are For: Benchmarking Spatial-Functional Intelligence in Multimodal LLMs","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07296","citing_title":"OpenSpatial: A Principled Data Engine for Empowering Spatial Intelligence","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2505.07062","citing_title":"Seed1.5-VL Technical Report","ref_index":164,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN","json":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN.json","graph_json":"https://pith.science/api/pith-number/WSX4VCGEDWO5IKJJIMT5JSJKNN/graph.json","events_json":"https://pith.science/api/pith-number/WSX4VCGEDWO5IKJJIMT5JSJKNN/events.json","paper":"https://pith.science/paper/WSX4VCGE"},"agent_actions":{"view_html":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN","download_json":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN.json","view_paper":"https://pith.science/paper/WSX4VCGE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.15280&json=true","fetch_graph":"https://pith.science/api/pith-number/WSX4VCGEDWO5IKJJIMT5JSJKNN/graph.json","fetch_events":"https://pith.science/api/pith-number/WSX4VCGEDWO5IKJJIMT5JSJKNN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN/action/storage_attestation","attest_author":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN/action/author_attestation","sign_citation":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN/action/citation_signature","submit_replication":"https://pith.science/pith/WSX4VCGEDWO5IKJJIMT5JSJKNN/action/replication_record"}},"created_at":"2026-07-05T10:54:47.277506+00:00","updated_at":"2026-07-05T10:54:47.277506+00:00"}