{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:L6TSQPPR7QQKC7KNME7THGRULU","short_pith_number":"pith:L6TSQPPR","schema_version":"1.0","canonical_sha256":"5fa7283df1fc20a17d4d613f339a345d0061e44454b1884dafdee93c52616ec4","source":{"kind":"arxiv","id":"2606.02380","version":1},"attestation_state":"computed","paper":{"title":"SPADE-Bench: Evaluating Spontaneous Strategic Deception in Agents via Plan-Action Divergence","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Dong, Haowei Li, Jiaming Ji, Juntao Dai, Kaiyue Yang, Qirui Zheng, Wenxin Li, Yaodong Yang, Yingshui Tan, Yuyan Bu","submitted_at":"2026-06-01T15:28:34Z","abstract_excerpt":"As LLM-based agents expand their operational scope, reliability becomes a prerequisite for real-world deployment. However, in practical applications, human users cannot monitor every immediate behavior; instead, the execution process often remains a black box, leaving users dependent solely on the agent's self-reported updates. This opacity creates a critical risk: agents may present observer-facing reports that diverge from their executed actions, rendering the system uncontrollable, especially in high-stakes autonomous scenarios. We term such self-reported plan-action divergence as agent dec"},"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":"2606.02380","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-01T15:28:34Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c6382efb7ab96c1df0fb38416cd90fde1116883315c0077e2e4c4874fcc0d731","abstract_canon_sha256":"850dd26206d4007c4da0d4d99ab09c58ba5ce6b0f562cd42a344bf95cdfef496"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T03:04:57.675389Z","signature_b64":"YB8eLd9aMM4iF1H3gG1O8nxsxKJbG9h2cbjKtl0o+z9fv292FvYncD2+TqjVCj1q6osVtyKN0tObtiP5ojvoCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5fa7283df1fc20a17d4d613f339a345d0061e44454b1884dafdee93c52616ec4","last_reissued_at":"2026-06-02T03:04:57.674989Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T03:04:57.674989Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SPADE-Bench: Evaluating Spontaneous Strategic Deception in Agents via Plan-Action Divergence","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Bowen Dong, Haowei Li, Jiaming Ji, Juntao Dai, Kaiyue Yang, Qirui Zheng, Wenxin Li, Yaodong Yang, Yingshui Tan, Yuyan Bu","submitted_at":"2026-06-01T15:28:34Z","abstract_excerpt":"As LLM-based agents expand their operational scope, reliability becomes a prerequisite for real-world deployment. However, in practical applications, human users cannot monitor every immediate behavior; instead, the execution process often remains a black box, leaving users dependent solely on the agent's self-reported updates. This opacity creates a critical risk: agents may present observer-facing reports that diverge from their executed actions, rendering the system uncontrollable, especially in high-stakes autonomous scenarios. We term such self-reported plan-action divergence as agent dec"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.02380","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/2606.02380/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":"2606.02380","created_at":"2026-06-02T03:04:57.675044+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.02380v1","created_at":"2026-06-02T03:04:57.675044+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.02380","created_at":"2026-06-02T03:04:57.675044+00:00"},{"alias_kind":"pith_short_12","alias_value":"L6TSQPPR7QQK","created_at":"2026-06-02T03:04:57.675044+00:00"},{"alias_kind":"pith_short_16","alias_value":"L6TSQPPR7QQKC7KN","created_at":"2026-06-02T03:04:57.675044+00:00"},{"alias_kind":"pith_short_8","alias_value":"L6TSQPPR","created_at":"2026-06-02T03:04:57.675044+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/L6TSQPPR7QQKC7KNME7THGRULU","json":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU.json","graph_json":"https://pith.science/api/pith-number/L6TSQPPR7QQKC7KNME7THGRULU/graph.json","events_json":"https://pith.science/api/pith-number/L6TSQPPR7QQKC7KNME7THGRULU/events.json","paper":"https://pith.science/paper/L6TSQPPR"},"agent_actions":{"view_html":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU","download_json":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU.json","view_paper":"https://pith.science/paper/L6TSQPPR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.02380&json=true","fetch_graph":"https://pith.science/api/pith-number/L6TSQPPR7QQKC7KNME7THGRULU/graph.json","fetch_events":"https://pith.science/api/pith-number/L6TSQPPR7QQKC7KNME7THGRULU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU/action/storage_attestation","attest_author":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU/action/author_attestation","sign_citation":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU/action/citation_signature","submit_replication":"https://pith.science/pith/L6TSQPPR7QQKC7KNME7THGRULU/action/replication_record"}},"created_at":"2026-06-02T03:04:57.675044+00:00","updated_at":"2026-06-02T03:04:57.675044+00:00"}