{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:TQEG37VT7VG37SRF3OUGHU4U4B","short_pith_number":"pith:TQEG37VT","canonical_record":{"source":{"id":"2504.11837","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:52:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"36ebbd404c227fccd0f24ad4c620129d4d6617b6b1c7a8573a3c97a832f71a68","abstract_canon_sha256":"460144836836ad4e1d44466fc9179fc0447f93c09b85522c32bb005b6991d905"},"schema_version":"1.0"},"canonical_sha256":"9c086dfeb3fd4dbfca25dba863d394e05056a5f265a1dab42c35ed2e60ec632c","source":{"kind":"arxiv","id":"2504.11837","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.11837","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"arxiv_version","alias_value":"2504.11837v3","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11837","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_12","alias_value":"TQEG37VT7VG3","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_16","alias_value":"TQEG37VT7VG37SRF","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_8","alias_value":"TQEG37VT","created_at":"2026-06-19T16:09:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:TQEG37VT7VG37SRF3OUGHU4U4B","target":"record","payload":{"canonical_record":{"source":{"id":"2504.11837","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:52:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"36ebbd404c227fccd0f24ad4c620129d4d6617b6b1c7a8573a3c97a832f71a68","abstract_canon_sha256":"460144836836ad4e1d44466fc9179fc0447f93c09b85522c32bb005b6991d905"},"schema_version":"1.0"},"canonical_sha256":"9c086dfeb3fd4dbfca25dba863d394e05056a5f265a1dab42c35ed2e60ec632c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:09:48.467439Z","signature_b64":"OWccqrM9WaJY0FrCkqPrQMScZxhuV7YraJ5lpkQ/8RLsIqLQu5ZgQLEFDE5wt1Oj7y45RGiEWqSNCeYd9qXCDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9c086dfeb3fd4dbfca25dba863d394e05056a5f265a1dab42c35ed2e60ec632c","last_reissued_at":"2026-06-19T16:09:48.467019Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:09:48.467019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.11837","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":"2026-06-19T16:09:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/tJcA7NcqqkoZWvEYQ5RsVJemMRjonC02Wxx2a/fc6UANGmH5bIkIHvu2sHHkqfY9HUq0uqLgg1VJLSFozeNCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T22:04:55.784489Z"},"content_sha256":"46472e11147889a2e5237d9e8432d77294f39d160ae05c00938820fdd4cdab32","schema_version":"1.0","event_id":"sha256:46472e11147889a2e5237d9e8432d77294f39d160ae05c00938820fdd4cdab32"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:TQEG37VT7VG37SRF3OUGHU4U4B","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"EmoFSM: A Finite State Machine for Emotional Support Conversation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hongyan Li, Luo Ji, Qingqing Gu, Teng Chen, Xiaoyu Wang, Yong Chen, Yue Zhao, Zhonglin Jiang","submitted_at":"2025-04-16T07:52:06Z","abstract_excerpt":"Emotional support conversation (ESC) aims to alleviate people's emotional distress through effective conversations. Although large language models (LLMs) have made remarkable progress in ESC, most of these studies may not define the diagram from a state-model perspective, thereby providing a suboptimal solution for long-term satisfaction. To address such an issue, we leverage the Finite State Machine (FSM) on LLMs, and propose a framework called EmoFSM. Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy, and the final "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy and the final response upon each conversational turn. Substantial experiments on ESC datasets suggest that FiSMiness outperforms many baselines, including direct inference, self-refine, chain of thought, finetuning, and external-assisted methods, even those with many more parameters.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The premise that conversations defined from a state model perspective are required for optimal long-term satisfaction, and that a single LLM can reliably implement the full FSM cycle (emotion reasoning, strategy choice, response) without external modules or post-hoc fixes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"FiSMiness uses finite state machines on LLMs to structure emotional support conversations and claims better performance than direct inference, CoT, finetuning, and external methods on ESC datasets.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"82385f5c6ac2a84c3dd14d925123ac06c004620410279a904ffa48ec3cca178f"},"source":{"id":"2504.11837","kind":"arxiv","version":3},"verdict":{"id":"978509ad-3b3f-48bc-844c-b00630e8846c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-22T20:53:58.929047Z","strongest_claim":"Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy and the final response upon each conversational turn. Substantial experiments on ESC datasets suggest that FiSMiness outperforms many baselines, including direct inference, self-refine, chain of thought, finetuning, and external-assisted methods, even those with many more parameters.","one_line_summary":"FiSMiness uses finite state machines on LLMs to structure emotional support conversations and claims better performance than direct inference, CoT, finetuning, and external methods on ESC datasets.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The premise that conversations defined from a state model perspective are required for optimal long-term satisfaction, and that a single LLM can reliably implement the full FSM cycle (emotion reasoning, strategy choice, response) without external modules or post-hoc fixes.","pith_extraction_headline":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2504.11837/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":"978509ad-3b3f-48bc-844c-b00630e8846c"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-19T16:09:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EmU3kOCUhXrl50Xl55NxmuxtZByMAaLD8E2vC1TZYt8zI+RUbDRWWaSDK0zU47RbGG8UaZhyErmMw4wj2dZCBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T22:04:55.785762Z"},"content_sha256":"e25fa5836f2bf9b4f5cd3e010eb06e8152360ef370d1a49d4ab9bfb9e500bb6f","schema_version":"1.0","event_id":"sha256:e25fa5836f2bf9b4f5cd3e010eb06e8152360ef370d1a49d4ab9bfb9e500bb6f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TQEG37VT7VG37SRF3OUGHU4U4B/bundle.json","state_url":"https://pith.science/pith/TQEG37VT7VG37SRF3OUGHU4U4B/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TQEG37VT7VG37SRF3OUGHU4U4B/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":"2026-08-04T22:04:55Z","links":{"resolver":"https://pith.science/pith/TQEG37VT7VG37SRF3OUGHU4U4B","bundle":"https://pith.science/pith/TQEG37VT7VG37SRF3OUGHU4U4B/bundle.json","state":"https://pith.science/pith/TQEG37VT7VG37SRF3OUGHU4U4B/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TQEG37VT7VG37SRF3OUGHU4U4B/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TQEG37VT7VG37SRF3OUGHU4U4B","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":"460144836836ad4e1d44466fc9179fc0447f93c09b85522c32bb005b6991d905","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:52:06Z","title_canon_sha256":"36ebbd404c227fccd0f24ad4c620129d4d6617b6b1c7a8573a3c97a832f71a68"},"schema_version":"1.0","source":{"id":"2504.11837","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.11837","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"arxiv_version","alias_value":"2504.11837v3","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11837","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_12","alias_value":"TQEG37VT7VG3","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_16","alias_value":"TQEG37VT7VG37SRF","created_at":"2026-06-19T16:09:48Z"},{"alias_kind":"pith_short_8","alias_value":"TQEG37VT","created_at":"2026-06-19T16:09:48Z"}],"graph_snapshots":[{"event_id":"sha256:e25fa5836f2bf9b4f5cd3e010eb06e8152360ef370d1a49d4ab9bfb9e500bb6f","target":"graph","created_at":"2026-06-19T16:09:48Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy and the final response upon each conversational turn. Substantial experiments on ESC datasets suggest that FiSMiness outperforms many baselines, including direct inference, self-refine, chain of thought, finetuning, and external-assisted methods, even those with many more parameters."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The premise that conversations defined from a state model perspective are required for optimal long-term satisfaction, and that a single LLM can reliably implement the full FSM cycle (emotion reasoning, strategy choice, response) without external modules or post-hoc fixes."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"FiSMiness uses finite state machines on LLMs to structure emotional support conversations and claims better performance than direct inference, CoT, finetuning, and external methods on ESC datasets."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn."}],"snapshot_sha256":"82385f5c6ac2a84c3dd14d925123ac06c004620410279a904ffa48ec3cca178f"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2504.11837/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Emotional support conversation (ESC) aims to alleviate people's emotional distress through effective conversations. Although large language models (LLMs) have made remarkable progress in ESC, most of these studies may not define the diagram from a state-model perspective, thereby providing a suboptimal solution for long-term satisfaction. To address such an issue, we leverage the Finite State Machine (FSM) on LLMs, and propose a framework called EmoFSM. Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy, and the final ","authors_text":"Hongyan Li, Luo Ji, Qingqing Gu, Teng Chen, Xiaoyu Wang, Yong Chen, Yue Zhao, Zhonglin Jiang","cross_cats":["cs.AI"],"headline":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn.","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:52:06Z","title":"EmoFSM: A Finite State Machine for Emotional Support Conversation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11837","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-22T20:53:58.929047Z","id":"978509ad-3b3f-48bc-844c-b00630e8846c","model_set":{"reader":"grok-4.3"},"one_line_summary":"FiSMiness uses finite state machines on LLMs to structure emotional support conversations and claims better performance than direct inference, CoT, finetuning, and external methods on ESC datasets.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Modeling emotional support conversations as a finite state machine enables a single LLM to handle emotion reasoning, strategy selection, and response generation turn by turn.","strongest_claim":"Our framework allows a single LLM to bootstrap the planning during ESC, and self-reason the seeker's emotion, support strategy and the final response upon each conversational turn. Substantial experiments on ESC datasets suggest that FiSMiness outperforms many baselines, including direct inference, self-refine, chain of thought, finetuning, and external-assisted methods, even those with many more parameters.","weakest_assumption":"The premise that conversations defined from a state model perspective are required for optimal long-term satisfaction, and that a single LLM can reliably implement the full FSM cycle (emotion reasoning, strategy choice, response) without external modules or post-hoc fixes."}},"verdict_id":"978509ad-3b3f-48bc-844c-b00630e8846c"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:46472e11147889a2e5237d9e8432d77294f39d160ae05c00938820fdd4cdab32","target":"record","created_at":"2026-06-19T16:09:48Z","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":"460144836836ad4e1d44466fc9179fc0447f93c09b85522c32bb005b6991d905","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:52:06Z","title_canon_sha256":"36ebbd404c227fccd0f24ad4c620129d4d6617b6b1c7a8573a3c97a832f71a68"},"schema_version":"1.0","source":{"id":"2504.11837","kind":"arxiv","version":3}},"canonical_sha256":"9c086dfeb3fd4dbfca25dba863d394e05056a5f265a1dab42c35ed2e60ec632c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9c086dfeb3fd4dbfca25dba863d394e05056a5f265a1dab42c35ed2e60ec632c","first_computed_at":"2026-06-19T16:09:48.467019Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:09:48.467019Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OWccqrM9WaJY0FrCkqPrQMScZxhuV7YraJ5lpkQ/8RLsIqLQu5ZgQLEFDE5wt1Oj7y45RGiEWqSNCeYd9qXCDw==","signature_status":"signed_v1","signed_at":"2026-06-19T16:09:48.467439Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.11837","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:46472e11147889a2e5237d9e8432d77294f39d160ae05c00938820fdd4cdab32","sha256:e25fa5836f2bf9b4f5cd3e010eb06e8152360ef370d1a49d4ab9bfb9e500bb6f"],"state_sha256":"24a78bf93f338880de18f59d3d4237af4fda0b639938180c102d2c4d125f58ca"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UT1lHoaYYAlaTKNff7zoiI2CJ7xpeK+YtW7/4plmn8s0TrhhWixPZYYnLXrJKlg35GDUgq+AeEOhqWjtNeXuBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T22:04:55.793540Z","bundle_sha256":"2cf5bdc5f21fc6acf4e509a63d19c43f5f461ff789d10ea80a23bc4fd8356a49"}}