{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:DAQE7THIAHWEIJEUPA6M4NXLBO","short_pith_number":"pith:DAQE7THI","canonical_record":{"source":{"id":"2502.08503","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T15:34:45Z","cross_cats_sorted":[],"title_canon_sha256":"7f268c35acc1e4999740da75a1858eb8d322c1786debcb74aac663896670835a","abstract_canon_sha256":"a6be2300055de702d1a6e8a654a4b5ba3358e0be4d5e60c08e33d8c9806677ae"},"schema_version":"1.0"},"canonical_sha256":"18204fcce801ec442494783cce36eb0b99b2f08ce8753bce2b5571f7fc887d95","source":{"kind":"arxiv","id":"2502.08503","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.08503","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.08503v3","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08503","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_12","alias_value":"DAQE7THIAHWE","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_16","alias_value":"DAQE7THIAHWEIJEU","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_8","alias_value":"DAQE7THI","created_at":"2026-07-05T11:16:54Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:DAQE7THIAHWEIJEUPA6M4NXLBO","target":"record","payload":{"canonical_record":{"source":{"id":"2502.08503","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T15:34:45Z","cross_cats_sorted":[],"title_canon_sha256":"7f268c35acc1e4999740da75a1858eb8d322c1786debcb74aac663896670835a","abstract_canon_sha256":"a6be2300055de702d1a6e8a654a4b5ba3358e0be4d5e60c08e33d8c9806677ae"},"schema_version":"1.0"},"canonical_sha256":"18204fcce801ec442494783cce36eb0b99b2f08ce8753bce2b5571f7fc887d95","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:54.877848Z","signature_b64":"VNT6/BEyFmwkCXtKB7LrcAWKGfKRxWTF3POqy7QHDJv7mPmdT+Q19FJqxMUdu+d1++prUV7Ohxzm9vNx4mNHAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18204fcce801ec442494783cce36eb0b99b2f08ce8753bce2b5571f7fc887d95","last_reissued_at":"2026-07-05T11:16:54.877385Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:54.877385Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.08503","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:16:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gqEZOOTT7njM1BwNhzy+dfnxGsxyaS14mRBYACVgpYm0+dJRgGJIfFYw1wEwYZikiVNJH1B9QXGEuHRmXhTTDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T01:29:47.649792Z"},"content_sha256":"51da78a0240a37f374428bd1b9733804c79e2dc8404d4083bea72653e5d9f096","schema_version":"1.0","event_id":"sha256:51da78a0240a37f374428bd1b9733804c79e2dc8404d4083bea72653e5d9f096"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:DAQE7THIAHWEIJEUPA6M4NXLBO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Revisiting 3D LLM Benchmarks: Are We Really Testing 3D Capabilities?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiahe Jin, Mingyan Yang, Yanheng He","submitted_at":"2025-02-12T15:34:45Z","abstract_excerpt":"In this work, we identify the \"2D-Cheating\" problem in 3D LLM evaluation, where these tasks might be easily solved by VLMs with rendered images of point clouds, exposing ineffective evaluation of 3D LLMs' unique 3D capabilities. We test VLM performance across multiple 3D LLM benchmarks and, using this as a reference, propose principles for better assessing genuine 3D understanding. We also advocate explicitly separating 3D abilities from 1D or 2D aspects when evaluating 3D LLMs. Code and data are available at https://github.com/LLM-class-group/Revisiting-3D-LLM-Benchmarks"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08503","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/2502.08503/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:16:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FpUS4ni9dLUwEtna6ek41inRfjJeviCVs6O+fLGqU0ip5Yq5iiim33guG41QH+IULuzi2lvYQkDRQG7zbVn/Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T01:29:47.650321Z"},"content_sha256":"a83295ba87647b8f3d48cdcf77f2834abe790b513ffdb4ca756ad24eba68a984","schema_version":"1.0","event_id":"sha256:a83295ba87647b8f3d48cdcf77f2834abe790b513ffdb4ca756ad24eba68a984"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/bundle.json","state_url":"https://pith.science/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-14T01:29:47Z","links":{"resolver":"https://pith.science/pith/DAQE7THIAHWEIJEUPA6M4NXLBO","bundle":"https://pith.science/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/bundle.json","state":"https://pith.science/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DAQE7THIAHWEIJEUPA6M4NXLBO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DAQE7THIAHWEIJEUPA6M4NXLBO","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"a6be2300055de702d1a6e8a654a4b5ba3358e0be4d5e60c08e33d8c9806677ae","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T15:34:45Z","title_canon_sha256":"7f268c35acc1e4999740da75a1858eb8d322c1786debcb74aac663896670835a"},"schema_version":"1.0","source":{"id":"2502.08503","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.08503","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"arxiv_version","alias_value":"2502.08503v3","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08503","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_12","alias_value":"DAQE7THIAHWE","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_16","alias_value":"DAQE7THIAHWEIJEU","created_at":"2026-07-05T11:16:54Z"},{"alias_kind":"pith_short_8","alias_value":"DAQE7THI","created_at":"2026-07-05T11:16:54Z"}],"graph_snapshots":[{"event_id":"sha256:a83295ba87647b8f3d48cdcf77f2834abe790b513ffdb4ca756ad24eba68a984","target":"graph","created_at":"2026-07-05T11:16:54Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.08503/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this work, we identify the \"2D-Cheating\" problem in 3D LLM evaluation, where these tasks might be easily solved by VLMs with rendered images of point clouds, exposing ineffective evaluation of 3D LLMs' unique 3D capabilities. We test VLM performance across multiple 3D LLM benchmarks and, using this as a reference, propose principles for better assessing genuine 3D understanding. We also advocate explicitly separating 3D abilities from 1D or 2D aspects when evaluating 3D LLMs. Code and data are available at https://github.com/LLM-class-group/Revisiting-3D-LLM-Benchmarks","authors_text":"Jiahe Jin, Mingyan Yang, Yanheng He","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T15:34:45Z","title":"Revisiting 3D LLM Benchmarks: Are We Really Testing 3D Capabilities?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08503","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:51da78a0240a37f374428bd1b9733804c79e2dc8404d4083bea72653e5d9f096","target":"record","created_at":"2026-07-05T11:16:54Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"a6be2300055de702d1a6e8a654a4b5ba3358e0be4d5e60c08e33d8c9806677ae","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-02-12T15:34:45Z","title_canon_sha256":"7f268c35acc1e4999740da75a1858eb8d322c1786debcb74aac663896670835a"},"schema_version":"1.0","source":{"id":"2502.08503","kind":"arxiv","version":3}},"canonical_sha256":"18204fcce801ec442494783cce36eb0b99b2f08ce8753bce2b5571f7fc887d95","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"18204fcce801ec442494783cce36eb0b99b2f08ce8753bce2b5571f7fc887d95","first_computed_at":"2026-07-05T11:16:54.877385Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:16:54.877385Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VNT6/BEyFmwkCXtKB7LrcAWKGfKRxWTF3POqy7QHDJv7mPmdT+Q19FJqxMUdu+d1++prUV7Ohxzm9vNx4mNHAg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:16:54.877848Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.08503","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:51da78a0240a37f374428bd1b9733804c79e2dc8404d4083bea72653e5d9f096","sha256:a83295ba87647b8f3d48cdcf77f2834abe790b513ffdb4ca756ad24eba68a984"],"state_sha256":"3596cc48fdfa3a7ac86e65ccf16dd5fd0aa01513ec230b0fd06e47b05c45ac88"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PdNcW3L/2BqDVNhI8LzLoDN2u1I5P2eYAXN5LIF+7tGDoE3YDcycEpNepSUj4UUNS3AdQX9XKG0KnGZrZTR9AA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T01:29:47.669036Z","bundle_sha256":"ccd3f46e65af6054161caba81daf79baf21641058cb192ec35b32cc83863e425"}}