{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:E44EJ52FFPTOVF7SSLFGRS6I3G","short_pith_number":"pith:E44EJ52F","canonical_record":{"source":{"id":"2507.16572","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-22T13:24:42Z","cross_cats_sorted":[],"title_canon_sha256":"0195dc4e44ff6a6cbbf3ba3c6e27f46cacd8a3efddb039d441bb8e9eaf3d2a5f","abstract_canon_sha256":"507da66bf032375646cbb6ed12726889bb0e395c8664ab26c64c0d611f59cf45"},"schema_version":"1.0"},"canonical_sha256":"273844f7452be6ea97f292ca68cbc8d9865cc7876d1b06fbd4154330e8b8ec2b","source":{"kind":"arxiv","id":"2507.16572","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.16572","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"arxiv_version","alias_value":"2507.16572v1","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.16572","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_12","alias_value":"E44EJ52FFPTO","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_16","alias_value":"E44EJ52FFPTOVF7S","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_8","alias_value":"E44EJ52F","created_at":"2026-07-05T11:41:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:E44EJ52FFPTOVF7SSLFGRS6I3G","target":"record","payload":{"canonical_record":{"source":{"id":"2507.16572","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-22T13:24:42Z","cross_cats_sorted":[],"title_canon_sha256":"0195dc4e44ff6a6cbbf3ba3c6e27f46cacd8a3efddb039d441bb8e9eaf3d2a5f","abstract_canon_sha256":"507da66bf032375646cbb6ed12726889bb0e395c8664ab26c64c0d611f59cf45"},"schema_version":"1.0"},"canonical_sha256":"273844f7452be6ea97f292ca68cbc8d9865cc7876d1b06fbd4154330e8b8ec2b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:41:12.327901Z","signature_b64":"t14BhDSDQv05noGzxr43ZWWE8vzdIM/obIlZ/nNpdEIDaJ6Fw8F+USKyvi/1ZDPAi3o6E9xYBWpi/AvG3JMgDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"273844f7452be6ea97f292ca68cbc8d9865cc7876d1b06fbd4154330e8b8ec2b","last_reissued_at":"2026-07-05T11:41:12.327444Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:41:12.327444Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.16572","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:41:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OXcnIhxtSIdt0Fx3GELRbn7hik1nu/YQnREg/HdE1sOv3S1ndbgQsgwFEk97Jgyh971meBZUiBtDlaeZQN+RCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T23:27:27.272273Z"},"content_sha256":"305845665d414d7df31c0caa2209104713bd47ead82f39c6479f14eb60d47f8b","schema_version":"1.0","event_id":"sha256:305845665d414d7df31c0caa2209104713bd47ead82f39c6479f14eb60d47f8b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:E44EJ52FFPTOVF7SSLFGRS6I3G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Pixels to Principles: Probing Intuitive Physics Understanding in Multimodal Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Elia Bruni, Mohamad Ballout, Serwan Jassim","submitted_at":"2025-07-22T13:24:42Z","abstract_excerpt":"This paper presents a systematic evaluation of state-of-the-art multimodal large language models (MLLMs) on intuitive physics tasks using the GRASP and IntPhys 2 datasets. We assess the open-source models InternVL 2.5, Qwen 2.5 VL, LLaVA-OneVision, and the proprietary Gemini 2.0 Flash Thinking, finding that even the latest models struggle to reliably distinguish physically plausible from implausible scenarios. To go beyond performance metrics, we conduct a probing analysis of model embeddings, extracting intermediate representations at key processing stages to examine how well task-relevant in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.16572","kind":"arxiv","version":1},"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/2507.16572/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":"2026-07-05T11:41:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"F8wYAY8izvXuBVDsZwpFKLTDOQxmB+dfGX29P3NzbIHkHNEMvATvB4BJXdDGlwS9mpKT0eVGh2JnE+FutziSBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T23:27:27.272764Z"},"content_sha256":"51952ccc1b069316dc60d676e0b552c5ad0de70be1b1213b248441022e05fe40","schema_version":"1.0","event_id":"sha256:51952ccc1b069316dc60d676e0b552c5ad0de70be1b1213b248441022e05fe40"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/bundle.json","state_url":"https://pith.science/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/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":"2026-08-07T23:27:27Z","links":{"resolver":"https://pith.science/pith/E44EJ52FFPTOVF7SSLFGRS6I3G","bundle":"https://pith.science/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/bundle.json","state":"https://pith.science/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/E44EJ52FFPTOVF7SSLFGRS6I3G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:E44EJ52FFPTOVF7SSLFGRS6I3G","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":"507da66bf032375646cbb6ed12726889bb0e395c8664ab26c64c0d611f59cf45","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-22T13:24:42Z","title_canon_sha256":"0195dc4e44ff6a6cbbf3ba3c6e27f46cacd8a3efddb039d441bb8e9eaf3d2a5f"},"schema_version":"1.0","source":{"id":"2507.16572","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.16572","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"arxiv_version","alias_value":"2507.16572v1","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.16572","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_12","alias_value":"E44EJ52FFPTO","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_16","alias_value":"E44EJ52FFPTOVF7S","created_at":"2026-07-05T11:41:12Z"},{"alias_kind":"pith_short_8","alias_value":"E44EJ52F","created_at":"2026-07-05T11:41:12Z"}],"graph_snapshots":[{"event_id":"sha256:51952ccc1b069316dc60d676e0b552c5ad0de70be1b1213b248441022e05fe40","target":"graph","created_at":"2026-07-05T11:41:12Z","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/2507.16572/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper presents a systematic evaluation of state-of-the-art multimodal large language models (MLLMs) on intuitive physics tasks using the GRASP and IntPhys 2 datasets. We assess the open-source models InternVL 2.5, Qwen 2.5 VL, LLaVA-OneVision, and the proprietary Gemini 2.0 Flash Thinking, finding that even the latest models struggle to reliably distinguish physically plausible from implausible scenarios. To go beyond performance metrics, we conduct a probing analysis of model embeddings, extracting intermediate representations at key processing stages to examine how well task-relevant in","authors_text":"Elia Bruni, Mohamad Ballout, Serwan Jassim","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-22T13:24:42Z","title":"Pixels to Principles: Probing Intuitive Physics Understanding in Multimodal Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.16572","kind":"arxiv","version":1},"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":"sha256:305845665d414d7df31c0caa2209104713bd47ead82f39c6479f14eb60d47f8b","target":"record","created_at":"2026-07-05T11:41:12Z","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":"507da66bf032375646cbb6ed12726889bb0e395c8664ab26c64c0d611f59cf45","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-22T13:24:42Z","title_canon_sha256":"0195dc4e44ff6a6cbbf3ba3c6e27f46cacd8a3efddb039d441bb8e9eaf3d2a5f"},"schema_version":"1.0","source":{"id":"2507.16572","kind":"arxiv","version":1}},"canonical_sha256":"273844f7452be6ea97f292ca68cbc8d9865cc7876d1b06fbd4154330e8b8ec2b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"273844f7452be6ea97f292ca68cbc8d9865cc7876d1b06fbd4154330e8b8ec2b","first_computed_at":"2026-07-05T11:41:12.327444Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:41:12.327444Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"t14BhDSDQv05noGzxr43ZWWE8vzdIM/obIlZ/nNpdEIDaJ6Fw8F+USKyvi/1ZDPAi3o6E9xYBWpi/AvG3JMgDg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:41:12.327901Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.16572","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:305845665d414d7df31c0caa2209104713bd47ead82f39c6479f14eb60d47f8b","sha256:51952ccc1b069316dc60d676e0b552c5ad0de70be1b1213b248441022e05fe40"],"state_sha256":"d5f7ee82c2c3f9d2342f724f1f88a81c0c57779dafade9f5fd700d668d516be7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mqT6zLPIMwlikPfhoTPHFjCphb12/J0xirDB5ZBcbsHaUZoH86V5uS319DaluBEdlSWaLbz7eN5BwJOfhqNmAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T23:27:27.278235Z","bundle_sha256":"db129ba0ef51b156e47c30678266a730d42f02c85d62a0d95d98aa3f014bde89"}}