{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3LA5NQTWPM2KGAW363MI52MBWQ","short_pith_number":"pith:3LA5NQTW","canonical_record":{"source":{"id":"2507.15521","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-21T11:42:03Z","cross_cats_sorted":[],"title_canon_sha256":"cb038cca266e1eca84a337de8164397a58f71bcd84891ae79f36ac89e0eecf5e","abstract_canon_sha256":"d40898e1b5302fdc6e44c5cbf3a7531e3257d403ee3f28f7029a7b2d81d3c241"},"schema_version":"1.0"},"canonical_sha256":"dac1d6c2767b34a302dbf6d88ee981b42f6b5f2b6a39a526fd2f1bcdff8e7ae5","source":{"kind":"arxiv","id":"2507.15521","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.15521","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"arxiv_version","alias_value":"2507.15521v1","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.15521","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_12","alias_value":"3LA5NQTWPM2K","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_16","alias_value":"3LA5NQTWPM2KGAW3","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_8","alias_value":"3LA5NQTW","created_at":"2026-07-05T11:40:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3LA5NQTWPM2KGAW363MI52MBWQ","target":"record","payload":{"canonical_record":{"source":{"id":"2507.15521","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-21T11:42:03Z","cross_cats_sorted":[],"title_canon_sha256":"cb038cca266e1eca84a337de8164397a58f71bcd84891ae79f36ac89e0eecf5e","abstract_canon_sha256":"d40898e1b5302fdc6e44c5cbf3a7531e3257d403ee3f28f7029a7b2d81d3c241"},"schema_version":"1.0"},"canonical_sha256":"dac1d6c2767b34a302dbf6d88ee981b42f6b5f2b6a39a526fd2f1bcdff8e7ae5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:40:38.664962Z","signature_b64":"PAGMezPPH3oEgCsuXEpZ9omTAl9Cy5ecIuleYdS+0l5u6beh1ndmbQaCKHStDwNAzPGY6RSlfzVKK7T51AhCCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dac1d6c2767b34a302dbf6d88ee981b42f6b5f2b6a39a526fd2f1bcdff8e7ae5","last_reissued_at":"2026-07-05T11:40:38.664581Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:40:38.664581Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.15521","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":"2026-07-05T11:40:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"47JtJzFYeatrTUsJa2SrrOJt6hbZFcMDfeGcj/mUiEosGmhEJyD4YvHxk8p9uLHG7XXI5/GtQpIeuCr+/OJmCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T09:40:20.660264Z"},"content_sha256":"434f959907de99aed88f24231cb697107f18dee7e77883519807e8eb39061eb3","schema_version":"1.0","event_id":"sha256:434f959907de99aed88f24231cb697107f18dee7e77883519807e8eb39061eb3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3LA5NQTWPM2KGAW363MI52MBWQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLM world models are mental: Output layer evidence of brittle world model use in LLM mechanical reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Cole Robertson, Philip Wolff","submitted_at":"2025-07-21T11:42:03Z","abstract_excerpt":"Do large language models (LLMs) construct and manipulate internal world models, or do they rely solely on statistical associations represented as output layer token probabilities? We adapt cognitive science methodologies from human mental models research to test LLMs on pulley system problems using TikZ-rendered stimuli. Study 1 examines whether LLMs can estimate mechanical advantage (MA). State-of-the-art models performed marginally but significantly above chance, and their estimates correlated significantly with ground-truth MA. Significant correlations between number of pulleys and model es"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.15521","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/2507.15521/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":"2026-07-05T11:40:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DdP60/e8+aQopNYTTk9W44Pf28bLPW/7xTYgfT+6MKhVqgZWMuiq3iP40WKiOpkjqMbmUx86nKJI8JpXUhZjBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T09:40:20.661047Z"},"content_sha256":"55d1945cfaabf2488bd638c9a09aa4042a3346857d33e210d6e3b63919cf2883","schema_version":"1.0","event_id":"sha256:55d1945cfaabf2488bd638c9a09aa4042a3346857d33e210d6e3b63919cf2883"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3LA5NQTWPM2KGAW363MI52MBWQ/bundle.json","state_url":"https://pith.science/pith/3LA5NQTWPM2KGAW363MI52MBWQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3LA5NQTWPM2KGAW363MI52MBWQ/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":"2026-08-08T09:40:20Z","links":{"resolver":"https://pith.science/pith/3LA5NQTWPM2KGAW363MI52MBWQ","bundle":"https://pith.science/pith/3LA5NQTWPM2KGAW363MI52MBWQ/bundle.json","state":"https://pith.science/pith/3LA5NQTWPM2KGAW363MI52MBWQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3LA5NQTWPM2KGAW363MI52MBWQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3LA5NQTWPM2KGAW363MI52MBWQ","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":"d40898e1b5302fdc6e44c5cbf3a7531e3257d403ee3f28f7029a7b2d81d3c241","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-21T11:42:03Z","title_canon_sha256":"cb038cca266e1eca84a337de8164397a58f71bcd84891ae79f36ac89e0eecf5e"},"schema_version":"1.0","source":{"id":"2507.15521","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.15521","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"arxiv_version","alias_value":"2507.15521v1","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.15521","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_12","alias_value":"3LA5NQTWPM2K","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_16","alias_value":"3LA5NQTWPM2KGAW3","created_at":"2026-07-05T11:40:38Z"},{"alias_kind":"pith_short_8","alias_value":"3LA5NQTW","created_at":"2026-07-05T11:40:38Z"}],"graph_snapshots":[{"event_id":"sha256:55d1945cfaabf2488bd638c9a09aa4042a3346857d33e210d6e3b63919cf2883","target":"graph","created_at":"2026-07-05T11:40:38Z","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/2507.15521/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Do large language models (LLMs) construct and manipulate internal world models, or do they rely solely on statistical associations represented as output layer token probabilities? We adapt cognitive science methodologies from human mental models research to test LLMs on pulley system problems using TikZ-rendered stimuli. Study 1 examines whether LLMs can estimate mechanical advantage (MA). State-of-the-art models performed marginally but significantly above chance, and their estimates correlated significantly with ground-truth MA. Significant correlations between number of pulleys and model es","authors_text":"Cole Robertson, Philip Wolff","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-21T11:42:03Z","title":"LLM world models are mental: Output layer evidence of brittle world model use in LLM mechanical reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.15521","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":"sha256:434f959907de99aed88f24231cb697107f18dee7e77883519807e8eb39061eb3","target":"record","created_at":"2026-07-05T11:40:38Z","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":"d40898e1b5302fdc6e44c5cbf3a7531e3257d403ee3f28f7029a7b2d81d3c241","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-07-21T11:42:03Z","title_canon_sha256":"cb038cca266e1eca84a337de8164397a58f71bcd84891ae79f36ac89e0eecf5e"},"schema_version":"1.0","source":{"id":"2507.15521","kind":"arxiv","version":1}},"canonical_sha256":"dac1d6c2767b34a302dbf6d88ee981b42f6b5f2b6a39a526fd2f1bcdff8e7ae5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dac1d6c2767b34a302dbf6d88ee981b42f6b5f2b6a39a526fd2f1bcdff8e7ae5","first_computed_at":"2026-07-05T11:40:38.664581Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:40:38.664581Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PAGMezPPH3oEgCsuXEpZ9omTAl9Cy5ecIuleYdS+0l5u6beh1ndmbQaCKHStDwNAzPGY6RSlfzVKK7T51AhCCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:40:38.664962Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.15521","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:434f959907de99aed88f24231cb697107f18dee7e77883519807e8eb39061eb3","sha256:55d1945cfaabf2488bd638c9a09aa4042a3346857d33e210d6e3b63919cf2883"],"state_sha256":"be5daf5ee3a4ecdb8dfb4c597a53e2a2d4caa51bc939e6aba121161552a07526"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jttPxmWYjC5RGOjbbNDzh93OYh4AKHJN0U/+uVStg4wcfUD7O40UBYmPPA4iQVC0WFIRoQR638GhcE6tRfOFBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T09:40:20.666283Z","bundle_sha256":"7878a18c3d6127f2d9bb038641a4e2a1de0509854dc94fbbec30d265bd8f3de1"}}