{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RXOMZM7RRNZLYYBN7ALVDAVRFI","short_pith_number":"pith:RXOMZM7R","canonical_record":{"source":{"id":"2605.12651","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T18:57:25Z","cross_cats_sorted":[],"title_canon_sha256":"6aa738ea0e118a7f178499a3d14148b5c6bc10f84787040fb9cf4024f1e6070c","abstract_canon_sha256":"7bf049c27be3d7c8fa82c492a22d90556f5966053e087ce9c2a620a174a36c3e"},"schema_version":"1.0"},"canonical_sha256":"8ddcccb3f18b72bc602df8175182b12a1614e0a8407ba24a254deb387c157967","source":{"kind":"arxiv","id":"2605.12651","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12651","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12651v2","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12651","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"pith_short_12","alias_value":"RXOMZM7RRNZL","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RXOMZM7RRNZLYYBN","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RXOMZM7R","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RXOMZM7RRNZLYYBN7ALVDAVRFI","target":"record","payload":{"canonical_record":{"source":{"id":"2605.12651","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T18:57:25Z","cross_cats_sorted":[],"title_canon_sha256":"6aa738ea0e118a7f178499a3d14148b5c6bc10f84787040fb9cf4024f1e6070c","abstract_canon_sha256":"7bf049c27be3d7c8fa82c492a22d90556f5966053e087ce9c2a620a174a36c3e"},"schema_version":"1.0"},"canonical_sha256":"8ddcccb3f18b72bc602df8175182b12a1614e0a8407ba24a254deb387c157967","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:09:50.788486Z","signature_b64":"c0yd6HGA/9fMy9La5W0pFlqIQGjg7GNFjd7A6y1pQvnXPK1g6BaKP4mPY3+WDBYXRaoCBH0UYpTQ0zsUM70GAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ddcccb3f18b72bc602df8175182b12a1614e0a8407ba24a254deb387c157967","last_reissued_at":"2026-05-18T03:09:50.787606Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:09:50.787606Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.12651","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:09:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"F6pHfRSG9B3dlNPScK68GRiZXjG2lgHLYaD5GO4j6nwUFSYTd5TeOHWkLHgIQXv3kw+owUAIfXUXiUk26LPWCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T13:57:14.759359Z"},"content_sha256":"2c9f6bb932d45077629bb5189d1695ad2dce30424b0d7551811916b009895d51","schema_version":"1.0","event_id":"sha256:2c9f6bb932d45077629bb5189d1695ad2dce30424b0d7551811916b009895d51"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RXOMZM7RRNZLYYBN7ALVDAVRFI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Runtime Monitoring of Perception-Based Autonomous Systems via Embedding Temporal Logic","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Abigail Hammer, Ashish Kapoor, Eunsuk Kang, Karen Leung, Parv Kapoor","submitted_at":"2026-05-12T18:57:25Z","abstract_excerpt":"Runtime monitoring of autonomous systems traditionally relies on mapping continuous sensor observations to discrete logical propositions defined over low-dimensional state variables. This abstraction breaks down in perception-driven settings, where such mappings require additional learned modules that are often computationally expensive, brittle, and semantically misaligned. In this work, we propose Embedding Temporal Logic (ETL), a temporal logic that performs monitoring directly in learned embedding spaces. ETL defines predicates through distances between observed embeddings and target embed"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ETL defines predicates through distances between observed embeddings and target embeddings derived from reference observations, allowing specifications to capture high-level perceptual concepts such as similarity to visual goals or avoidance of semantic regions.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That distances in the learned embedding space reliably and semantically align with the perceptual concepts of interest for the predicates to be meaningful and that conformal calibration provides safety-oriented reliability without additional assumptions on the embedding model.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Embedding Temporal Logic (ETL) performs runtime monitoring directly in learned embedding spaces using distance-based predicates composed with temporal operators, supported by conformal calibration for reliable predicate evaluation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4919c792d16d5c90a316ab40bac360e04c03f679534c893f50eaf1e9bbe2771d"},"source":{"id":"2605.12651","kind":"arxiv","version":2},"verdict":{"id":"6bc0f178-40d0-435d-8592-ea12760e6d94","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:00:04.080641Z","strongest_claim":"ETL defines predicates through distances between observed embeddings and target embeddings derived from reference observations, allowing specifications to capture high-level perceptual concepts such as similarity to visual goals or avoidance of semantic regions.","one_line_summary":"Embedding Temporal Logic (ETL) performs runtime monitoring directly in learned embedding spaces using distance-based predicates composed with temporal operators, supported by conformal calibration for reliable predicate evaluation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That distances in the learned embedding space reliably and semantically align with the perceptual concepts of interest for the predicates to be meaningful and that conformal calibration provides safety-oriented reliability without additional assumptions on the embedding model.","pith_extraction_headline":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings."},"references":{"count":140,"sample":[{"doi":"","year":2024,"title":"2024 IEEE International Conference on Robotics and Automation (ICRA) , pages=","work_id":"9e09b93a-a6a0-440a-80de-c9ecde256b8e","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1063/1.5024914","year":null,"title":"Hauke and Donner, Reik V","work_id":"cffbf648-8d1b-4303-914f-47c6b59072cc","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Lang2LTL-2: Grounding Spatiotemporal Navigation Commands Using Large Language and Vision-Language Models , year=","work_id":"d63a11ca-bb3f-4d2b-b1cb-a9a94e5c2c0d","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"AnySafe: Adapting Latent Safety Filters at Runtime via Safety Constraint Parameterization in the Latent Space , author=. 2025 , eprint=","work_id":"75e923c2-59a9-47e2-aab5-de0b8c779586","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"2021 , howpublished =","work_id":"04c7ecd6-c9ef-4e95-b7d7-eaf799c93ad1","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":140,"snapshot_sha256":"9df8c3c2183eaba2a6611c132b4825315ab6656eef6eacbe872073ce36221a62","internal_anchors":12},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ada6e48cd45e3ea4749e4256eaa89265f7e6ca473cad6e5ada502842393fd87f"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"6bc0f178-40d0-435d-8592-ea12760e6d94"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:09:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ekFXkEXtseosOh3nMmPIVFuDME5tcRiQ1KIOi7Qkk5umcNv94pheAuieMznELzTJCXv50Plu2dx2/Edb9XWXBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T13:57:14.759883Z"},"content_sha256":"695a16d6dd717b961c46eda3eb151372a1045e0de1175b5b2b9d4947f5ca0eee","schema_version":"1.0","event_id":"sha256:695a16d6dd717b961c46eda3eb151372a1045e0de1175b5b2b9d4947f5ca0eee"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/bundle.json","state_url":"https://pith.science/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/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":"2026-06-02T13:57:14Z","links":{"resolver":"https://pith.science/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI","bundle":"https://pith.science/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/bundle.json","state":"https://pith.science/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RXOMZM7RRNZLYYBN7ALVDAVRFI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RXOMZM7RRNZLYYBN7ALVDAVRFI","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":"7bf049c27be3d7c8fa82c492a22d90556f5966053e087ce9c2a620a174a36c3e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T18:57:25Z","title_canon_sha256":"6aa738ea0e118a7f178499a3d14148b5c6bc10f84787040fb9cf4024f1e6070c"},"schema_version":"1.0","source":{"id":"2605.12651","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12651","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12651v2","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12651","created_at":"2026-05-18T03:09:50Z"},{"alias_kind":"pith_short_12","alias_value":"RXOMZM7RRNZL","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RXOMZM7RRNZLYYBN","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RXOMZM7R","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:695a16d6dd717b961c46eda3eb151372a1045e0de1175b5b2b9d4947f5ca0eee","target":"graph","created_at":"2026-05-18T03:09:50Z","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":"ETL defines predicates through distances between observed embeddings and target embeddings derived from reference observations, allowing specifications to capture high-level perceptual concepts such as similarity to visual goals or avoidance of semantic regions."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That distances in the learned embedding space reliably and semantically align with the perceptual concepts of interest for the predicates to be meaningful and that conformal calibration provides safety-oriented reliability without additional assumptions on the embedding model."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Embedding Temporal Logic (ETL) performs runtime monitoring directly in learned embedding spaces using distance-based predicates composed with temporal operators, supported by conformal calibration for reliable predicate evaluation."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings."}],"snapshot_sha256":"4919c792d16d5c90a316ab40bac360e04c03f679534c893f50eaf1e9bbe2771d"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ada6e48cd45e3ea4749e4256eaa89265f7e6ca473cad6e5ada502842393fd87f"},"paper":{"abstract_excerpt":"Runtime monitoring of autonomous systems traditionally relies on mapping continuous sensor observations to discrete logical propositions defined over low-dimensional state variables. This abstraction breaks down in perception-driven settings, where such mappings require additional learned modules that are often computationally expensive, brittle, and semantically misaligned. In this work, we propose Embedding Temporal Logic (ETL), a temporal logic that performs monitoring directly in learned embedding spaces. ETL defines predicates through distances between observed embeddings and target embed","authors_text":"Abigail Hammer, Ashish Kapoor, Eunsuk Kang, Karen Leung, Parv Kapoor","cross_cats":[],"headline":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T18:57:25Z","title":"Runtime Monitoring of Perception-Based Autonomous Systems via Embedding Temporal Logic"},"references":{"count":140,"internal_anchors":12,"resolved_work":140,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"2024 IEEE International Conference on Robotics and Automation (ICRA) , pages=","work_id":"9e09b93a-a6a0-440a-80de-c9ecde256b8e","year":2024},{"cited_arxiv_id":"","doi":"10.1063/1.5024914","is_internal_anchor":false,"ref_index":2,"title":"Hauke and Donner, Reik V","work_id":"cffbf648-8d1b-4303-914f-47c6b59072cc","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Lang2LTL-2: Grounding Spatiotemporal Navigation Commands Using Large Language and Vision-Language Models , year=","work_id":"d63a11ca-bb3f-4d2b-b1cb-a9a94e5c2c0d","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"AnySafe: Adapting Latent Safety Filters at Runtime via Safety Constraint Parameterization in the Latent Space , author=. 2025 , eprint=","work_id":"75e923c2-59a9-47e2-aab5-de0b8c779586","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"2021 , howpublished =","work_id":"04c7ecd6-c9ef-4e95-b7d7-eaf799c93ad1","year":2021}],"snapshot_sha256":"9df8c3c2183eaba2a6611c132b4825315ab6656eef6eacbe872073ce36221a62"},"source":{"id":"2605.12651","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-15T05:00:04.080641Z","id":"6bc0f178-40d0-435d-8592-ea12760e6d94","model_set":{"reader":"grok-4.3"},"one_line_summary":"Embedding Temporal Logic (ETL) performs runtime monitoring directly in learned embedding spaces using distance-based predicates composed with temporal operators, supported by conformal calibration for reliable predicate evaluation.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Embedding Temporal Logic monitors autonomous systems by defining predicates on distances between observed and reference embeddings.","strongest_claim":"ETL defines predicates through distances between observed embeddings and target embeddings derived from reference observations, allowing specifications to capture high-level perceptual concepts such as similarity to visual goals or avoidance of semantic regions.","weakest_assumption":"That distances in the learned embedding space reliably and semantically align with the perceptual concepts of interest for the predicates to be meaningful and that conformal calibration provides safety-oriented reliability without additional assumptions on the embedding model."}},"verdict_id":"6bc0f178-40d0-435d-8592-ea12760e6d94"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2c9f6bb932d45077629bb5189d1695ad2dce30424b0d7551811916b009895d51","target":"record","created_at":"2026-05-18T03:09:50Z","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":"7bf049c27be3d7c8fa82c492a22d90556f5966053e087ce9c2a620a174a36c3e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T18:57:25Z","title_canon_sha256":"6aa738ea0e118a7f178499a3d14148b5c6bc10f84787040fb9cf4024f1e6070c"},"schema_version":"1.0","source":{"id":"2605.12651","kind":"arxiv","version":2}},"canonical_sha256":"8ddcccb3f18b72bc602df8175182b12a1614e0a8407ba24a254deb387c157967","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8ddcccb3f18b72bc602df8175182b12a1614e0a8407ba24a254deb387c157967","first_computed_at":"2026-05-18T03:09:50.787606Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:09:50.787606Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"c0yd6HGA/9fMy9La5W0pFlqIQGjg7GNFjd7A6y1pQvnXPK1g6BaKP4mPY3+WDBYXRaoCBH0UYpTQ0zsUM70GAg==","signature_status":"signed_v1","signed_at":"2026-05-18T03:09:50.788486Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.12651","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2c9f6bb932d45077629bb5189d1695ad2dce30424b0d7551811916b009895d51","sha256:695a16d6dd717b961c46eda3eb151372a1045e0de1175b5b2b9d4947f5ca0eee"],"state_sha256":"a426cd4bf017974af033ff762064dcf2e5312f4caa800ac182cef7cec1e9a288"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fdtshyTaw5pJP/0A6oluiLq9kvP8g9DtmMMZTJIW8j3gGYJQEpejdpKThyO/bgqhBqDpzoyPhyVPItJImp0KDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T13:57:14.762230Z","bundle_sha256":"fd94038f92fc87fb1ffca58669af8aed5ad45bca11e2bc7c8f2c1e9d583ebf0d"}}