{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:UVNIMJFAC6XWGHUR2IJ2AQ75M3","short_pith_number":"pith:UVNIMJFA","canonical_record":{"source":{"id":"2310.15421","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-24T00:24:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7dc246af7c875d31b6ee7c177b5a24a1b37e039f0a056be02ad08d1bc7ce9abf","abstract_canon_sha256":"63ff32bd1f5cb01253a72f000e8448501e95177ec5d82d2a0d2df49cdd61911f"},"schema_version":"1.0"},"canonical_sha256":"a55a8624a017af631e91d213a043fd66ea44b019cae145b27f8b96348a635c30","source":{"kind":"arxiv","id":"2310.15421","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.15421","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2310.15421v3","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.15421","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"UVNIMJFAC6XW","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"UVNIMJFAC6XWGHUR","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"UVNIMJFA","created_at":"2026-07-05T07:07:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:UVNIMJFAC6XWGHUR2IJ2AQ75M3","target":"record","payload":{"canonical_record":{"source":{"id":"2310.15421","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-24T00:24:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7dc246af7c875d31b6ee7c177b5a24a1b37e039f0a056be02ad08d1bc7ce9abf","abstract_canon_sha256":"63ff32bd1f5cb01253a72f000e8448501e95177ec5d82d2a0d2df49cdd61911f"},"schema_version":"1.0"},"canonical_sha256":"a55a8624a017af631e91d213a043fd66ea44b019cae145b27f8b96348a635c30","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:07:18.254963Z","signature_b64":"DHTlE2Guc5+AY3K25ykMrG1t9nyEgqXQhBn71sSSVpRAEpIuhidRipzeCw7qES+J8azgbPsb1t9D72fuz0BhBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a55a8624a017af631e91d213a043fd66ea44b019cae145b27f8b96348a635c30","last_reissued_at":"2026-07-05T07:07:18.254490Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:07:18.254490Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.15421","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":"2026-07-05T07:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zrafZ2jD1VP1IQI18oFCRAtILE1012TSdLKM5XVkWcAWECpYr/mfZD+9rsfSlW4iBNstj26lQvAhbSJF17RTBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T21:04:35.510004Z"},"content_sha256":"53ab9380e323b622ea4748ae4d29ce869e266a36ffbd03f90e3d75643ce0edc8","schema_version":"1.0","event_id":"sha256:53ab9380e323b622ea4748ae4d29ce869e266a36ffbd03f90e3d75643ce0edc8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:UVNIMJFAC6XWGHUR2IJ2AQ75M3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Gunhee Kim, Hyunwoo Kim, Maarten Sap, Melanie Sclar, Ronan Le Bras, Xuhui Zhou, Yejin Choi","submitted_at":"2023-10-24T00:24:11Z","abstract_excerpt":"Theory of mind (ToM) evaluations currently focus on testing models using passive narratives that inherently lack interactivity. We introduce FANToM, a new benchmark designed to stress-test ToM within information-asymmetric conversational contexts via question answering. Our benchmark draws upon important theoretical requisites from psychology and necessary empirical considerations when evaluating large language models (LLMs). In particular, we formulate multiple types of questions that demand the same underlying reasoning to identify illusory or false sense of ToM capabilities in LLMs. We show"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.15421","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/2310.15421/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":"2026-07-05T07:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oKERnCMQGRKnl/f+1NGQTD7Dhf6s2U8TMM1wqs9t65I5V5uJR1FRa72Yw8K7bhkwJStywNtwuW4yLXbgK7uCAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T21:04:35.510503Z"},"content_sha256":"388ad5e6ba08e2f80c2fe71573c18799ad52514f0ccbb4b3b8db266b3dc16b54","schema_version":"1.0","event_id":"sha256:388ad5e6ba08e2f80c2fe71573c18799ad52514f0ccbb4b3b8db266b3dc16b54"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/bundle.json","state_url":"https://pith.science/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/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":"2026-08-09T21:04:35Z","links":{"resolver":"https://pith.science/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3","bundle":"https://pith.science/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/bundle.json","state":"https://pith.science/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/UVNIMJFAC6XWGHUR2IJ2AQ75M3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:UVNIMJFAC6XWGHUR2IJ2AQ75M3","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":"63ff32bd1f5cb01253a72f000e8448501e95177ec5d82d2a0d2df49cdd61911f","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-24T00:24:11Z","title_canon_sha256":"7dc246af7c875d31b6ee7c177b5a24a1b37e039f0a056be02ad08d1bc7ce9abf"},"schema_version":"1.0","source":{"id":"2310.15421","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.15421","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2310.15421v3","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.15421","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"UVNIMJFAC6XW","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"UVNIMJFAC6XWGHUR","created_at":"2026-07-05T07:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"UVNIMJFA","created_at":"2026-07-05T07:07:18Z"}],"graph_snapshots":[{"event_id":"sha256:388ad5e6ba08e2f80c2fe71573c18799ad52514f0ccbb4b3b8db266b3dc16b54","target":"graph","created_at":"2026-07-05T07:07:18Z","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/2310.15421/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Theory of mind (ToM) evaluations currently focus on testing models using passive narratives that inherently lack interactivity. We introduce FANToM, a new benchmark designed to stress-test ToM within information-asymmetric conversational contexts via question answering. Our benchmark draws upon important theoretical requisites from psychology and necessary empirical considerations when evaluating large language models (LLMs). In particular, we formulate multiple types of questions that demand the same underlying reasoning to identify illusory or false sense of ToM capabilities in LLMs. We show","authors_text":"Gunhee Kim, Hyunwoo Kim, Maarten Sap, Melanie Sclar, Ronan Le Bras, Xuhui Zhou, Yejin Choi","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-24T00:24:11Z","title":"FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.15421","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":"sha256:53ab9380e323b622ea4748ae4d29ce869e266a36ffbd03f90e3d75643ce0edc8","target":"record","created_at":"2026-07-05T07:07:18Z","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":"63ff32bd1f5cb01253a72f000e8448501e95177ec5d82d2a0d2df49cdd61911f","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-24T00:24:11Z","title_canon_sha256":"7dc246af7c875d31b6ee7c177b5a24a1b37e039f0a056be02ad08d1bc7ce9abf"},"schema_version":"1.0","source":{"id":"2310.15421","kind":"arxiv","version":3}},"canonical_sha256":"a55a8624a017af631e91d213a043fd66ea44b019cae145b27f8b96348a635c30","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a55a8624a017af631e91d213a043fd66ea44b019cae145b27f8b96348a635c30","first_computed_at":"2026-07-05T07:07:18.254490Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:07:18.254490Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DHTlE2Guc5+AY3K25ykMrG1t9nyEgqXQhBn71sSSVpRAEpIuhidRipzeCw7qES+J8azgbPsb1t9D72fuz0BhBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:07:18.254963Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.15421","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:53ab9380e323b622ea4748ae4d29ce869e266a36ffbd03f90e3d75643ce0edc8","sha256:388ad5e6ba08e2f80c2fe71573c18799ad52514f0ccbb4b3b8db266b3dc16b54"],"state_sha256":"7127907d66ca79a89b39fa287fd4aacd38f1653761b96ef46e75a3f0dc0acad1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QrvB05WpIVT+VZbxdQrevncLX9OLVXYB/JHeAznbAhck+PGZfIhortNJkAmNL4+Bgl8ispPQ9XeYc8MJu3YgCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T21:04:35.514616Z","bundle_sha256":"3b8b43f92cd9d66e757d116f134a2a37db657d2190339c450824dcf9505e5ea6"}}