{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KVF46EE7KUOUZV3VGOKSE32VET","short_pith_number":"pith:KVF46EE7","schema_version":"1.0","canonical_sha256":"554bcf109f551d4cd7753395226f5524d5733b96ed4f734733d85dff8617453a","source":{"kind":"arxiv","id":"2504.14650","version":1},"attestation_state":"computed","paper":{"title":"A Framework for Benchmarking and Aligning Task-Planning Safety in LLM-Based Embodied Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Leilei Ding, Mingxiao Ma, Tianfu Wang, Wuyang Zhang, Xinrui Lin, Yanyong Zhang, Yuting Huang, Zhipeng Tang","submitted_at":"2025-04-20T15:12:14Z","abstract_excerpt":"Large Language Models (LLMs) exhibit substantial promise in enhancing task-planning capabilities within embodied agents due to their advanced reasoning and comprehension. However, the systemic safety of these agents remains an underexplored frontier. In this study, we present Safe-BeAl, an integrated framework for the measurement (SafePlan-Bench) and alignment (Safe-Align) of LLM-based embodied agents' behaviors. SafePlan-Bench establishes a comprehensive benchmark for evaluating task-planning safety, encompassing 2,027 daily tasks and corresponding environments distributed across 8 distinct h"},"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":"2504.14650","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-04-20T15:12:14Z","cross_cats_sorted":[],"title_canon_sha256":"a06e8ad449f796e892eecfeec4a0e1ba0b15c0959d30b346973d821997cc6664","abstract_canon_sha256":"a7098694449a1a70f8f1cc0d7f6dec666e26f0c2e23b92f9658fe86536606d4e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:51:48.219560Z","signature_b64":"Qzm56xV9ks9BChmPalguCLOV9XEkh/Gr/1FoZ1Om6CJ9KS+E3iYTN/hXWwL1hJaVAkEggwxqX6e6uQi+VDiMCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"554bcf109f551d4cd7753395226f5524d5733b96ed4f734733d85dff8617453a","last_reissued_at":"2026-07-05T10:51:48.219012Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:51:48.219012Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Framework for Benchmarking and Aligning Task-Planning Safety in LLM-Based Embodied Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Leilei Ding, Mingxiao Ma, Tianfu Wang, Wuyang Zhang, Xinrui Lin, Yanyong Zhang, Yuting Huang, Zhipeng Tang","submitted_at":"2025-04-20T15:12:14Z","abstract_excerpt":"Large Language Models (LLMs) exhibit substantial promise in enhancing task-planning capabilities within embodied agents due to their advanced reasoning and comprehension. However, the systemic safety of these agents remains an underexplored frontier. In this study, we present Safe-BeAl, an integrated framework for the measurement (SafePlan-Bench) and alignment (Safe-Align) of LLM-based embodied agents' behaviors. SafePlan-Bench establishes a comprehensive benchmark for evaluating task-planning safety, encompassing 2,027 daily tasks and corresponding environments distributed across 8 distinct h"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.14650","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/2504.14650/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":"2504.14650","created_at":"2026-07-05T10:51:48.219075+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.14650v1","created_at":"2026-07-05T10:51:48.219075+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.14650","created_at":"2026-07-05T10:51:48.219075+00:00"},{"alias_kind":"pith_short_12","alias_value":"KVF46EE7KUOU","created_at":"2026-07-05T10:51:48.219075+00:00"},{"alias_kind":"pith_short_16","alias_value":"KVF46EE7KUOUZV3V","created_at":"2026-07-05T10:51:48.219075+00:00"},{"alias_kind":"pith_short_8","alias_value":"KVF46EE7","created_at":"2026-07-05T10:51:48.219075+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09038","citing_title":"Personalization Meets Safety:Mechanisms,Risks,and Mitigations in Personalized LLMs","ref_index":247,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02900","citing_title":"Safety in Embodied AI: A Survey of Risks, Attacks, and Defenses","ref_index":145,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19638","citing_title":"SafetyALFRED: Evaluating Safety-Conscious Planning of Multimodal Large Language Models","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18463","citing_title":"Using large language models for embodied planning introduces systematic safety risks","ref_index":6,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET","json":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET.json","graph_json":"https://pith.science/api/pith-number/KVF46EE7KUOUZV3VGOKSE32VET/graph.json","events_json":"https://pith.science/api/pith-number/KVF46EE7KUOUZV3VGOKSE32VET/events.json","paper":"https://pith.science/paper/KVF46EE7"},"agent_actions":{"view_html":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET","download_json":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET.json","view_paper":"https://pith.science/paper/KVF46EE7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.14650&json=true","fetch_graph":"https://pith.science/api/pith-number/KVF46EE7KUOUZV3VGOKSE32VET/graph.json","fetch_events":"https://pith.science/api/pith-number/KVF46EE7KUOUZV3VGOKSE32VET/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET/action/storage_attestation","attest_author":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET/action/author_attestation","sign_citation":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET/action/citation_signature","submit_replication":"https://pith.science/pith/KVF46EE7KUOUZV3VGOKSE32VET/action/replication_record"}},"created_at":"2026-07-05T10:51:48.219075+00:00","updated_at":"2026-07-05T10:51:48.219075+00:00"}