{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:I55WMNR7KBJM73RPHRW6JRC4IF","short_pith_number":"pith:I55WMNR7","canonical_record":{"source":{"id":"2501.08838","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-15T14:47:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fe839408f8d2e05a93110997d55c0f8f13f94fff878954d3bd19df38bcfe48e0","abstract_canon_sha256":"92b73cc77302a3925ed9ddb911352b983202ab239d954a24877169f7eb793cc6"},"schema_version":"1.0"},"canonical_sha256":"477b66363f5052cfee2f3c6de4c45c4149ad5d468c03050f2c2aea4c31bd6b5f","source":{"kind":"arxiv","id":"2501.08838","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.08838","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"arxiv_version","alias_value":"2501.08838v1","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.08838","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_12","alias_value":"I55WMNR7KBJM","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_16","alias_value":"I55WMNR7KBJM73RP","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_8","alias_value":"I55WMNR7","created_at":"2026-07-05T10:01:26Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:I55WMNR7KBJM73RPHRW6JRC4IF","target":"record","payload":{"canonical_record":{"source":{"id":"2501.08838","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-15T14:47:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fe839408f8d2e05a93110997d55c0f8f13f94fff878954d3bd19df38bcfe48e0","abstract_canon_sha256":"92b73cc77302a3925ed9ddb911352b983202ab239d954a24877169f7eb793cc6"},"schema_version":"1.0"},"canonical_sha256":"477b66363f5052cfee2f3c6de4c45c4149ad5d468c03050f2c2aea4c31bd6b5f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:01:26.910336Z","signature_b64":"WVREH5C38ItsTgoFLqKu9q1Dr81j1I/F138mJ7xopaC/ZLlXki8fWLv9WMg1lYRGdsbGK5Ym5nFmTJHuXDezCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"477b66363f5052cfee2f3c6de4c45c4149ad5d468c03050f2c2aea4c31bd6b5f","last_reissued_at":"2026-07-05T10:01:26.909897Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:01:26.909897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.08838","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":"2026-07-05T10:01:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EcHuyhWKh2sCXcwpDVAHLxMw3l59zkbkx3XOfHVn17OazDCiAQq7Xp6YcbBSN+9Yvwk83F3VhLG80Dvk/3NHAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T22:11:03.134940Z"},"content_sha256":"a533836b6e742edd6b5ba9240e92abbb02bba8610acfc8d37f96d42413f294ba","schema_version":"1.0","event_id":"sha256:a533836b6e742edd6b5ba9240e92abbb02bba8610acfc8d37f96d42413f294ba"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:I55WMNR7KBJM73RPHRW6JRC4IF","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ToMATO: Verbalizing the Mental States of Role-Playing LLMs for Benchmarking Theory of Mind","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hiroaki Sugiyama, Kazutoshi Shinoda, Keita Suzuki, Kuniko Saito, Kyosuke Nishida, Nobukatsu Hojo, Ryo Masumura, Saki Mizuno","submitted_at":"2025-01-15T14:47:02Z","abstract_excerpt":"Existing Theory of Mind (ToM) benchmarks diverge from real-world scenarios in three aspects: 1) they assess a limited range of mental states such as beliefs, 2) false beliefs are not comprehensively explored, and 3) the diverse personality traits of characters are overlooked. To address these challenges, we introduce ToMATO, a new ToM benchmark formulated as multiple-choice QA over conversations. ToMATO is generated via LLM-LLM conversations featuring information asymmetry. By employing a prompting method that requires role-playing LLMs to verbalize their thoughts before each utterance, we cap"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.08838","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/2501.08838/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":"2026-07-05T10:01:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rDAerTkcpZHxQ6NGUPUPFJNDFFKDe7XV9sa/wj/53TUFAb/5zfRd9U+/Ksx+TWnOnLbnnZirpcyFC9oS8FJwCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T22:11:03.135354Z"},"content_sha256":"94cec041cda599ef4ba00f29afc48e136133ce838ebf717c04d4c911f9827004","schema_version":"1.0","event_id":"sha256:94cec041cda599ef4ba00f29afc48e136133ce838ebf717c04d4c911f9827004"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/I55WMNR7KBJM73RPHRW6JRC4IF/bundle.json","state_url":"https://pith.science/pith/I55WMNR7KBJM73RPHRW6JRC4IF/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/I55WMNR7KBJM73RPHRW6JRC4IF/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":"2026-07-08T22:11:03Z","links":{"resolver":"https://pith.science/pith/I55WMNR7KBJM73RPHRW6JRC4IF","bundle":"https://pith.science/pith/I55WMNR7KBJM73RPHRW6JRC4IF/bundle.json","state":"https://pith.science/pith/I55WMNR7KBJM73RPHRW6JRC4IF/state.json","well_known_bundle":"https://pith.science/.well-known/pith/I55WMNR7KBJM73RPHRW6JRC4IF/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:I55WMNR7KBJM73RPHRW6JRC4IF","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":"92b73cc77302a3925ed9ddb911352b983202ab239d954a24877169f7eb793cc6","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-15T14:47:02Z","title_canon_sha256":"fe839408f8d2e05a93110997d55c0f8f13f94fff878954d3bd19df38bcfe48e0"},"schema_version":"1.0","source":{"id":"2501.08838","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.08838","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"arxiv_version","alias_value":"2501.08838v1","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.08838","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_12","alias_value":"I55WMNR7KBJM","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_16","alias_value":"I55WMNR7KBJM73RP","created_at":"2026-07-05T10:01:26Z"},{"alias_kind":"pith_short_8","alias_value":"I55WMNR7","created_at":"2026-07-05T10:01:26Z"}],"graph_snapshots":[{"event_id":"sha256:94cec041cda599ef4ba00f29afc48e136133ce838ebf717c04d4c911f9827004","target":"graph","created_at":"2026-07-05T10:01:26Z","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/2501.08838/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Existing Theory of Mind (ToM) benchmarks diverge from real-world scenarios in three aspects: 1) they assess a limited range of mental states such as beliefs, 2) false beliefs are not comprehensively explored, and 3) the diverse personality traits of characters are overlooked. To address these challenges, we introduce ToMATO, a new ToM benchmark formulated as multiple-choice QA over conversations. ToMATO is generated via LLM-LLM conversations featuring information asymmetry. By employing a prompting method that requires role-playing LLMs to verbalize their thoughts before each utterance, we cap","authors_text":"Hiroaki Sugiyama, Kazutoshi Shinoda, Keita Suzuki, Kuniko Saito, Kyosuke Nishida, Nobukatsu Hojo, Ryo Masumura, Saki Mizuno","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-15T14:47:02Z","title":"ToMATO: Verbalizing the Mental States of Role-Playing LLMs for Benchmarking Theory of Mind"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.08838","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:a533836b6e742edd6b5ba9240e92abbb02bba8610acfc8d37f96d42413f294ba","target":"record","created_at":"2026-07-05T10:01:26Z","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":"92b73cc77302a3925ed9ddb911352b983202ab239d954a24877169f7eb793cc6","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-15T14:47:02Z","title_canon_sha256":"fe839408f8d2e05a93110997d55c0f8f13f94fff878954d3bd19df38bcfe48e0"},"schema_version":"1.0","source":{"id":"2501.08838","kind":"arxiv","version":1}},"canonical_sha256":"477b66363f5052cfee2f3c6de4c45c4149ad5d468c03050f2c2aea4c31bd6b5f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"477b66363f5052cfee2f3c6de4c45c4149ad5d468c03050f2c2aea4c31bd6b5f","first_computed_at":"2026-07-05T10:01:26.909897Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:01:26.909897Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"WVREH5C38ItsTgoFLqKu9q1Dr81j1I/F138mJ7xopaC/ZLlXki8fWLv9WMg1lYRGdsbGK5Ym5nFmTJHuXDezCw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:01:26.910336Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.08838","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a533836b6e742edd6b5ba9240e92abbb02bba8610acfc8d37f96d42413f294ba","sha256:94cec041cda599ef4ba00f29afc48e136133ce838ebf717c04d4c911f9827004"],"state_sha256":"24e56615386650d07be2166a50a67c48496f594524ec9a7c6f3b5a3847d74316"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GHxhJx52P5NisyMUHDGBZA0lk0UyrvImjoJs05ofMHshwV8YDIhzGq5G+/ZvTZObIAcIHGdR6QNDwufZuyWxCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-08T22:11:03.137694Z","bundle_sha256":"08a8e8947186e85edeab3d8c43ffe7ab9744d36e1910ad678ba9279850377f93"}}