{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:RRWWA7SR75IM47GCA4PVSPZ44K","short_pith_number":"pith:RRWWA7SR","schema_version":"1.0","canonical_sha256":"8c6d607e51ff50ce7cc2071f593f3ce2b67ccc975cbd12fe76c8ec259def3530","source":{"kind":"arxiv","id":"2606.09092","version":1},"attestation_state":"computed","paper":{"title":"From Shortcuts to Reasoning: Robust Post-Training of Theory of Mind with Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Behzad Dariush, Jike Zhong, Konstantinos Psounis, Ming Li, Shao-Yuan Lo, Wuao Liu, Yuheng Li, Yuxiang Lai","submitted_at":"2026-06-08T06:42:12Z","abstract_excerpt":"Theory of Mind (ToM) is a must-acquire skill for modern foundation model systems to operate effectively and safely in the real world. Recent works have explored honing ToM via post-training; however, we show that such progress is confounded by a pervasive \"shortcut\" issue: tasks can reach up to 99% accuracy by simply exploiting spurious causal correlations, leading to a false sense of ToM. Motivated by this, we first develop a framework to systematically examine ToM datasets for shortcuts and provide guidance for future development. We find that questions reducible to pure state tracking, such"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.09092","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T06:42:12Z","cross_cats_sorted":[],"title_canon_sha256":"cb3e0cc0c583739f6a069ba987c64d32e56741bc1c1628e448382ca16d6f38a0","abstract_canon_sha256":"34c0eae91b2e2901894adc8a458477028d9073766ca3b9621e4cdb667d748e49"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:07:58.533620Z","signature_b64":"3cxsR1cnwQGC2araKGQBU1EzEDWG3c4PaolNW6omBR2ncD+a5UdZQ5gU7cJX2sl6svlxm241etDPb/p4qdG5BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8c6d607e51ff50ce7cc2071f593f3ce2b67ccc975cbd12fe76c8ec259def3530","last_reissued_at":"2026-06-09T02:07:58.532831Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:07:58.532831Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Shortcuts to Reasoning: Robust Post-Training of Theory of Mind with Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Behzad Dariush, Jike Zhong, Konstantinos Psounis, Ming Li, Shao-Yuan Lo, Wuao Liu, Yuheng Li, Yuxiang Lai","submitted_at":"2026-06-08T06:42:12Z","abstract_excerpt":"Theory of Mind (ToM) is a must-acquire skill for modern foundation model systems to operate effectively and safely in the real world. Recent works have explored honing ToM via post-training; however, we show that such progress is confounded by a pervasive \"shortcut\" issue: tasks can reach up to 99% accuracy by simply exploiting spurious causal correlations, leading to a false sense of ToM. Motivated by this, we first develop a framework to systematically examine ToM datasets for shortcuts and provide guidance for future development. We find that questions reducible to pure state tracking, such"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09092","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/2606.09092/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.09092","created_at":"2026-06-09T02:07:58.532934+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.09092v1","created_at":"2026-06-09T02:07:58.532934+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09092","created_at":"2026-06-09T02:07:58.532934+00:00"},{"alias_kind":"pith_short_12","alias_value":"RRWWA7SR75IM","created_at":"2026-06-09T02:07:58.532934+00:00"},{"alias_kind":"pith_short_16","alias_value":"RRWWA7SR75IM47GC","created_at":"2026-06-09T02:07:58.532934+00:00"},{"alias_kind":"pith_short_8","alias_value":"RRWWA7SR","created_at":"2026-06-09T02:07:58.532934+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K","json":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K.json","graph_json":"https://pith.science/api/pith-number/RRWWA7SR75IM47GCA4PVSPZ44K/graph.json","events_json":"https://pith.science/api/pith-number/RRWWA7SR75IM47GCA4PVSPZ44K/events.json","paper":"https://pith.science/paper/RRWWA7SR"},"agent_actions":{"view_html":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K","download_json":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K.json","view_paper":"https://pith.science/paper/RRWWA7SR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.09092&json=true","fetch_graph":"https://pith.science/api/pith-number/RRWWA7SR75IM47GCA4PVSPZ44K/graph.json","fetch_events":"https://pith.science/api/pith-number/RRWWA7SR75IM47GCA4PVSPZ44K/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K/action/storage_attestation","attest_author":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K/action/author_attestation","sign_citation":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K/action/citation_signature","submit_replication":"https://pith.science/pith/RRWWA7SR75IM47GCA4PVSPZ44K/action/replication_record"}},"created_at":"2026-06-09T02:07:58.532934+00:00","updated_at":"2026-06-09T02:07:58.532934+00:00"}