{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:F2CXBDEQ56ULNY2ZIN4A7UCG7C","short_pith_number":"pith:F2CXBDEQ","canonical_record":{"source":{"id":"2608.04646","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T10:05:53Z","cross_cats_sorted":[],"title_canon_sha256":"619d0563635a1e29f91672ee82ee7834ded2624386c4b53bd7fcb04e089870a0","abstract_canon_sha256":"98e6279861062f8ee696ef872ccbfe2c9f9678a222a2dc2f5da4da63abb0c166"},"schema_version":"1.0"},"canonical_sha256":"2e85708c90efa8b6e35943780fd046f8b6aa305f5d2db9b00911476060ac06d3","source":{"kind":"arxiv","id":"2608.04646","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04646","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04646v1","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04646","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_12","alias_value":"F2CXBDEQ56UL","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_16","alias_value":"F2CXBDEQ56ULNY2Z","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_8","alias_value":"F2CXBDEQ","created_at":"2026-08-06T01:36:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:F2CXBDEQ56ULNY2ZIN4A7UCG7C","target":"record","payload":{"canonical_record":{"source":{"id":"2608.04646","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T10:05:53Z","cross_cats_sorted":[],"title_canon_sha256":"619d0563635a1e29f91672ee82ee7834ded2624386c4b53bd7fcb04e089870a0","abstract_canon_sha256":"98e6279861062f8ee696ef872ccbfe2c9f9678a222a2dc2f5da4da63abb0c166"},"schema_version":"1.0"},"canonical_sha256":"2e85708c90efa8b6e35943780fd046f8b6aa305f5d2db9b00911476060ac06d3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:36:32.356216Z","signature_b64":"I4NlHSwBF1hQ+SOzN10NMM9COb0QTscAzmiO1B2hPFuTWmTwQeX7qfcOyN0a/KwoKQOtgUOHqI7IcNfwafpvAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2e85708c90efa8b6e35943780fd046f8b6aa305f5d2db9b00911476060ac06d3","last_reissued_at":"2026-08-06T01:36:32.354693Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:36:32.354693Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.04646","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":"2026-08-06T01:36:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BbeW+3PMA6ecjHns9sEeLwuG56FrmX/cuea0bKlJ0M4SJdGWqtsyZ8E5HcYUY0TDamaZH75wbn3iopJW4TPVBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T07:32:38.065606Z"},"content_sha256":"24f1aa49aa7a49a9a59a17b364d2ab78ffa051aae70ce1f7b0b44328398030e7","schema_version":"1.0","event_id":"sha256:24f1aa49aa7a49a9a59a17b364d2ab78ffa051aae70ce1f7b0b44328398030e7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:F2CXBDEQ56ULNY2ZIN4A7UCG7C","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Evaluating Theory of Mind in Reasoning Models: Robustness over Reasoning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ian B. de Haan, Max van Duijn, Peter van der Putten","submitted_at":"2026-08-05T10:05:53Z","abstract_excerpt":"Large language models (LLMs) have recently shown strong performance on Theory of Mind (ToM) tests, prompting debate about the nature and validity of the underlying capabilities. At the same time, reasoning-oriented LLMs trained via reinforcement learning with verifiable rewards have demonstrated notable improvements across a range of benchmarks. In this work, we examine the behavior of such reasoning models in ToM tasks using novel adaptations of machine psychological experiments together with results from established benchmarks. We observe that reasoning models consistently exhibit increased "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04646","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/2608.04646/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":"2026-08-06T01:36:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BdetOsVKXzHbtTNMT5c6UoruzYI47iKGs9nODHaCw3a/qWj0/i6ucC1uAtC/svrcuNULQDVhnpXcGCgMfYQ4Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T07:32:38.066765Z"},"content_sha256":"5b713ceb1580fd992f7cfad9c97e95b64a2639efc7a970c45a90bea73fa46766","schema_version":"1.0","event_id":"sha256:5b713ceb1580fd992f7cfad9c97e95b64a2639efc7a970c45a90bea73fa46766"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:F2CXBDEQ56ULNY2ZIN4A7UCG7C","target":"integrity","payload":{"note":"Identifier '10.1016/0010-0277(85' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","snippet":"Baron-Cohen, S., Leslie, A.M., Frith, U.: Does the autistic child have a \"theory of mind\"?Cognition21(1),37–46(Oct1985).https://doi.org/10.1016/0010-0277(85) 90022-8","arxiv_id":"2608.04646","detector":"doi_compliance","evidence":{"doi":"10.1016/0010-0277(85","arxiv_id":null,"ref_index":1,"raw_excerpt":"Baron-Cohen, S., Leslie, A.M., Frith, U.: Does the autistic child have a \"theory of mind\"?Cognition21(1),37–46(Oct1985).https://doi.org/10.1016/0010-0277(85) 90022-8","parse_status":"well_formed","verdict_class":"cross_source","checked_sources":["crossref_by_doi","openalex_by_doi","doi_org_head"],"resolution_status":"hard_miss"},"severity":"critical","ref_index":1,"audited_at":"2026-08-06T19:58:17.252891Z","event_type":"pith.integrity.v1","detected_doi":"10.1016/0010-0277(85","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"unresolvable_identifier","evidence_hash":"71621a90f0c85cb756f25e9efcad071328a0f628401a08ec04c5f8e5189fb99e","paper_version":1,"verdict_class":"cross_source","resolved_title":null,"detector_version":"1.1.0","detected_arxiv_id":null,"integrity_event_id":18356,"payload_sha256":"c80ad042e9bd805ee3e18c843d3cc90c228802ec3e0a26540429e8e001b40073","signature_b64":"M8vFlJrcYPRN1ShleQTOZe12MKHNd9cqEtknjIubAyN7tnFS17pBYTMTBCEVlzRmkN+xQUOw9Iza3b/bCAIWBg==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-06T19:58:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nrh9aW/XFqKs2/Yw3OD8HPNE75tRoLNd1RSS8znIle/+vww8suFEKtmNWlhq6ZILfaj9cehfbkMPJfoOoo97Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T07:32:38.073777Z"},"content_sha256":"a1f7d9cb38d0ff13e5c2b9b03b6d8fbe1ec1a073fa03c7a4ff56ad6f64e1332c","schema_version":"1.0","event_id":"sha256:a1f7d9cb38d0ff13e5c2b9b03b6d8fbe1ec1a073fa03c7a4ff56ad6f64e1332c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/bundle.json","state_url":"https://pith.science/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/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":"2026-08-10T07:32:38Z","links":{"resolver":"https://pith.science/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C","bundle":"https://pith.science/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/bundle.json","state":"https://pith.science/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/state.json","well_known_bundle":"https://pith.science/.well-known/pith/F2CXBDEQ56ULNY2ZIN4A7UCG7C/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:F2CXBDEQ56ULNY2ZIN4A7UCG7C","merge_version":"pith-open-graph-merge-v1","event_count":3,"valid_event_count":3,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"98e6279861062f8ee696ef872ccbfe2c9f9678a222a2dc2f5da4da63abb0c166","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T10:05:53Z","title_canon_sha256":"619d0563635a1e29f91672ee82ee7834ded2624386c4b53bd7fcb04e089870a0"},"schema_version":"1.0","source":{"id":"2608.04646","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04646","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04646v1","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04646","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_12","alias_value":"F2CXBDEQ56UL","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_16","alias_value":"F2CXBDEQ56ULNY2Z","created_at":"2026-08-06T01:36:32Z"},{"alias_kind":"pith_short_8","alias_value":"F2CXBDEQ","created_at":"2026-08-06T01:36:32Z"}],"graph_snapshots":[{"event_id":"sha256:5b713ceb1580fd992f7cfad9c97e95b64a2639efc7a970c45a90bea73fa46766","target":"graph","created_at":"2026-08-06T01:36:32Z","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/2608.04646/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have recently shown strong performance on Theory of Mind (ToM) tests, prompting debate about the nature and validity of the underlying capabilities. At the same time, reasoning-oriented LLMs trained via reinforcement learning with verifiable rewards have demonstrated notable improvements across a range of benchmarks. In this work, we examine the behavior of such reasoning models in ToM tasks using novel adaptations of machine psychological experiments together with results from established benchmarks. We observe that reasoning models consistently exhibit increased ","authors_text":"Ian B. de Haan, Max van Duijn, Peter van der Putten","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T10:05:53Z","title":"Evaluating Theory of Mind in Reasoning Models: Robustness over Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04646","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":"sha256:24f1aa49aa7a49a9a59a17b364d2ab78ffa051aae70ce1f7b0b44328398030e7","target":"record","created_at":"2026-08-06T01:36:32Z","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":"98e6279861062f8ee696ef872ccbfe2c9f9678a222a2dc2f5da4da63abb0c166","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T10:05:53Z","title_canon_sha256":"619d0563635a1e29f91672ee82ee7834ded2624386c4b53bd7fcb04e089870a0"},"schema_version":"1.0","source":{"id":"2608.04646","kind":"arxiv","version":1}},"canonical_sha256":"2e85708c90efa8b6e35943780fd046f8b6aa305f5d2db9b00911476060ac06d3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2e85708c90efa8b6e35943780fd046f8b6aa305f5d2db9b00911476060ac06d3","first_computed_at":"2026-08-06T01:36:32.354693Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-06T01:36:32.354693Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"I4NlHSwBF1hQ+SOzN10NMM9COb0QTscAzmiO1B2hPFuTWmTwQeX7qfcOyN0a/KwoKQOtgUOHqI7IcNfwafpvAQ==","signature_status":"signed_v1","signed_at":"2026-08-06T01:36:32.356216Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.04646","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:24f1aa49aa7a49a9a59a17b364d2ab78ffa051aae70ce1f7b0b44328398030e7","sha256:5b713ceb1580fd992f7cfad9c97e95b64a2639efc7a970c45a90bea73fa46766","sha256:a1f7d9cb38d0ff13e5c2b9b03b6d8fbe1ec1a073fa03c7a4ff56ad6f64e1332c"],"state_sha256":"a5491d2127312a88aab890022add59c41441bca18f96dd03478136f842429d08"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"DdojKf2DYPCeKcMGtVYjpHY+jUqAvLvTWgmZx/EfH2Tk/1jsgNLc9ZUgvDTewmoy6jNdXaVGRwt8VQVijLDfAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T07:32:38.077162Z","bundle_sha256":"0c5633b4a8bbcc3674125e75766c6130fed521d30674f7ed77a7d0de5be8d819"}}