{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:P5WE7NHA7HIF4AA55VLZHUSI7C","short_pith_number":"pith:P5WE7NHA","canonical_record":{"source":{"id":"2409.14986","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-23T13:05:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b2cb00e573c3a5000550764d47c2ffc00ad8084ced3da5cf7d7dd1731d4f6c2b","abstract_canon_sha256":"cd2f9a5d99cd0089d230e71a36794456cbdbbf4633ed15d1b30ce608d4c3aafb"},"schema_version":"1.0"},"canonical_sha256":"7f6c4fb4e0f9d05e001ded5793d248f891b47783b4431b9abe592eca409f1d3d","source":{"kind":"arxiv","id":"2409.14986","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.14986","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"arxiv_version","alias_value":"2409.14986v1","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.14986","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_12","alias_value":"P5WE7NHA7HIF","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_16","alias_value":"P5WE7NHA7HIF4AA5","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_8","alias_value":"P5WE7NHA","created_at":"2026-07-05T09:10:34Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:P5WE7NHA7HIF4AA55VLZHUSI7C","target":"record","payload":{"canonical_record":{"source":{"id":"2409.14986","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-23T13:05:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b2cb00e573c3a5000550764d47c2ffc00ad8084ced3da5cf7d7dd1731d4f6c2b","abstract_canon_sha256":"cd2f9a5d99cd0089d230e71a36794456cbdbbf4633ed15d1b30ce608d4c3aafb"},"schema_version":"1.0"},"canonical_sha256":"7f6c4fb4e0f9d05e001ded5793d248f891b47783b4431b9abe592eca409f1d3d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:10:34.326345Z","signature_b64":"eXWuHsvz2Y4qeA8qSbMuvZgFPJawom/BHXtaXKcIONLu0SCF2f2S6uAzdBV5rrEAymYWYwBrJaht9FAH76wOAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7f6c4fb4e0f9d05e001ded5793d248f891b47783b4431b9abe592eca409f1d3d","last_reissued_at":"2026-07-05T09:10:34.325857Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:10:34.325857Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.14986","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":"2026-07-05T09:10:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"saT/q1AqIs+9TQDvG4STp2Zy7W+aWv8MDUcgCHwL+4+IVQAKpELcYk5JHFh/46pbF+0NZjiLcwIrWGrwLLU1Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T09:59:11.989378Z"},"content_sha256":"82c7255f839eecd15489c47a4f34b572585ebd3a9343aebf611b4fb93c160ee8","schema_version":"1.0","event_id":"sha256:82c7255f839eecd15489c47a4f34b572585ebd3a9343aebf611b4fb93c160ee8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:P5WE7NHA7HIF4AA55VLZHUSI7C","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Evaluating Theory of (an uncertain) Mind: Predicting the Uncertain Beliefs of Others in Conversation Forecasting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Anthony Sicilia, Malihe Alikhani","submitted_at":"2024-09-23T13:05:25Z","abstract_excerpt":"Typically, when evaluating Theory of Mind, we consider the beliefs of others to be binary: held or not held. But what if someone is unsure about their own beliefs? How can we quantify this uncertainty? We propose a new suite of tasks, challenging language models (LMs) to model the uncertainty of others in dialogue. We design these tasks around conversation forecasting, wherein an agent forecasts an unobserved outcome to a conversation. Uniquely, we view interlocutors themselves as forecasters, asking an LM to predict the uncertainty of the interlocutors (a probability). We experiment with re-s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.14986","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/2409.14986/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":"2026-07-05T09:10:34Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jhySeLBKGwylsFncPY7xpTF1yMZxtwi0iiu1QIBrY5yJ5wqAMeL1yEhH/upBFTfHk6nuZ6n/2L8+ya89dI98Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T09:59:11.989961Z"},"content_sha256":"9af4d60bc48bf4e6ef82618351225f1f0ef16da431089712619a3021df26b1b4","schema_version":"1.0","event_id":"sha256:9af4d60bc48bf4e6ef82618351225f1f0ef16da431089712619a3021df26b1b4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/bundle.json","state_url":"https://pith.science/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/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":"2026-08-12T09:59:11Z","links":{"resolver":"https://pith.science/pith/P5WE7NHA7HIF4AA55VLZHUSI7C","bundle":"https://pith.science/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/bundle.json","state":"https://pith.science/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/state.json","well_known_bundle":"https://pith.science/.well-known/pith/P5WE7NHA7HIF4AA55VLZHUSI7C/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:P5WE7NHA7HIF4AA55VLZHUSI7C","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":"cd2f9a5d99cd0089d230e71a36794456cbdbbf4633ed15d1b30ce608d4c3aafb","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-23T13:05:25Z","title_canon_sha256":"b2cb00e573c3a5000550764d47c2ffc00ad8084ced3da5cf7d7dd1731d4f6c2b"},"schema_version":"1.0","source":{"id":"2409.14986","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.14986","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"arxiv_version","alias_value":"2409.14986v1","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.14986","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_12","alias_value":"P5WE7NHA7HIF","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_16","alias_value":"P5WE7NHA7HIF4AA5","created_at":"2026-07-05T09:10:34Z"},{"alias_kind":"pith_short_8","alias_value":"P5WE7NHA","created_at":"2026-07-05T09:10:34Z"}],"graph_snapshots":[{"event_id":"sha256:9af4d60bc48bf4e6ef82618351225f1f0ef16da431089712619a3021df26b1b4","target":"graph","created_at":"2026-07-05T09:10:34Z","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/2409.14986/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Typically, when evaluating Theory of Mind, we consider the beliefs of others to be binary: held or not held. But what if someone is unsure about their own beliefs? How can we quantify this uncertainty? We propose a new suite of tasks, challenging language models (LMs) to model the uncertainty of others in dialogue. We design these tasks around conversation forecasting, wherein an agent forecasts an unobserved outcome to a conversation. Uniquely, we view interlocutors themselves as forecasters, asking an LM to predict the uncertainty of the interlocutors (a probability). We experiment with re-s","authors_text":"Anthony Sicilia, Malihe Alikhani","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-23T13:05:25Z","title":"Evaluating Theory of (an uncertain) Mind: Predicting the Uncertain Beliefs of Others in Conversation Forecasting"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.14986","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":"sha256:82c7255f839eecd15489c47a4f34b572585ebd3a9343aebf611b4fb93c160ee8","target":"record","created_at":"2026-07-05T09:10:34Z","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":"cd2f9a5d99cd0089d230e71a36794456cbdbbf4633ed15d1b30ce608d4c3aafb","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-23T13:05:25Z","title_canon_sha256":"b2cb00e573c3a5000550764d47c2ffc00ad8084ced3da5cf7d7dd1731d4f6c2b"},"schema_version":"1.0","source":{"id":"2409.14986","kind":"arxiv","version":1}},"canonical_sha256":"7f6c4fb4e0f9d05e001ded5793d248f891b47783b4431b9abe592eca409f1d3d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7f6c4fb4e0f9d05e001ded5793d248f891b47783b4431b9abe592eca409f1d3d","first_computed_at":"2026-07-05T09:10:34.325857Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:10:34.325857Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eXWuHsvz2Y4qeA8qSbMuvZgFPJawom/BHXtaXKcIONLu0SCF2f2S6uAzdBV5rrEAymYWYwBrJaht9FAH76wOAg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:10:34.326345Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.14986","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:82c7255f839eecd15489c47a4f34b572585ebd3a9343aebf611b4fb93c160ee8","sha256:9af4d60bc48bf4e6ef82618351225f1f0ef16da431089712619a3021df26b1b4"],"state_sha256":"ff7aab8cc85b1c1060ec81fbcca1b7a61234249f08ef5bd2f4aa00c440e1ab38"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IWY7Zb4kr5fpxGUj/WJc0eopGAgS23nGw0pWZT6TyFlMN2Ld4eaJS8vFaOhRXHkYJPB/1YVbe61hrOSMYLnoAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-12T09:59:11.993685Z","bundle_sha256":"e30b7e22bf68eb77a75afcc7aaecfb9499dde0eaf0dca5d8ddfcb405bcf8db99"}}