{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:PEBP4TFO5XRL37FWVRS5A2Q35Y","short_pith_number":"pith:PEBP4TFO","canonical_record":{"source":{"id":"2510.18034","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-10-20T19:14:29Z","cross_cats_sorted":["cs.AI","cs.RO"],"title_canon_sha256":"aa1b55cc34fcf2b2d0ea87381c2ee6c38355f7435c3639f9b8b6cb07a037c177","abstract_canon_sha256":"593f05e84dc7ad1f8d210d0208c7e01fd913cc1d269b611aad76b7acca25ea21"},"schema_version":"1.0"},"canonical_sha256":"7902fe4caeede2bdfcb6ac65d06a1bee3be3190dc6b428d4b303493b606df8d6","source":{"kind":"arxiv","id":"2510.18034","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.18034","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"arxiv_version","alias_value":"2510.18034v3","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.18034","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_12","alias_value":"PEBP4TFO5XRL","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_16","alias_value":"PEBP4TFO5XRL37FW","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_8","alias_value":"PEBP4TFO","created_at":"2026-05-21T01:05:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:PEBP4TFO5XRL37FWVRS5A2Q35Y","target":"record","payload":{"canonical_record":{"source":{"id":"2510.18034","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-10-20T19:14:29Z","cross_cats_sorted":["cs.AI","cs.RO"],"title_canon_sha256":"aa1b55cc34fcf2b2d0ea87381c2ee6c38355f7435c3639f9b8b6cb07a037c177","abstract_canon_sha256":"593f05e84dc7ad1f8d210d0208c7e01fd913cc1d269b611aad76b7acca25ea21"},"schema_version":"1.0"},"canonical_sha256":"7902fe4caeede2bdfcb6ac65d06a1bee3be3190dc6b428d4b303493b606df8d6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:05:11.176306Z","signature_b64":"fKIjdb15PYV3eV5EhYl4CDt6FOTldCogDrFwF53tzZZ9Dd3/zV4IeqHXS0tpUtwD1pshCDiifvbH9Dkk8eZZAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7902fe4caeede2bdfcb6ac65d06a1bee3be3190dc6b428d4b303493b606df8d6","last_reissued_at":"2026-05-21T01:05:11.175417Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:05:11.175417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2510.18034","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":"2026-05-21T01:05:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"j+BpMg/7mkv7G66tK7jERG7CLrUgSPO2ytcJBueHWkbb7ENHLEZFTp5v0lVdVcA/rKl8SfgK1qefaXhN8czeDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T05:34:10.169211Z"},"content_sha256":"128c6f264134f31586bb21ce4b4c2b9c1991f2b74c894bb72fc6549d6665d2c7","schema_version":"1.0","event_id":"sha256:128c6f264134f31586bb21ce4b4c2b9c1991f2b74c894bb72fc6549d6665d2c7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:PEBP4TFO5XRL37FWVRS5A2Q35Y","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can VLMs Unlock Semantic Anomaly Detection? A Framework for Structured Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall.","cross_cats":["cs.AI","cs.RO"],"primary_cat":"cs.CV","authors_text":"David Pop, Johannes Betz, Mattia Piccinini, Roberto Brusnicki, Yuan Gao","submitted_at":"2025-10-20T19:14:29Z","abstract_excerpt":"Autonomous driving systems remain critically vulnerable to the long-tail of rare, out-of-distribution semantic anomalies. While VLMs have emerged as promising tools for perception, their application in anomaly detection remains largely restricted to prompting proprietary models - limiting reliability, reproducibility, and deployment feasibility. To address this gap, we introduce SAVANT (Semantic Anomaly Verification/Analysis Toolkit), a novel model-agnostic reasoning framework that reformulates anomaly detection as a layered semantic consistency verification. By applying SAVANT's two-phase pip"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Applying SAVANT improves VLM's absolute recall by approximately 18.5% compared to prompting baselines, and leveraging the best proprietary model within the framework enables automatic labeling of around 10,000 images to fine-tune a 7B open-source model achieving 90.8% recall and 93.8% accuracy.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The evaluation uses a 'balanced set of real-world driving scenarios' whose selection criteria and representativeness of long-tail anomalies are not specified, which is required to support that the reported recall gains are due to the structured reasoning pipeline rather than dataset construction.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SAVANT boosts VLM recall for semantic anomaly detection in driving images by 18.5% via structured reasoning and enables fine-tuning a 7B open model to 90.8% recall and 93.8% accuracy.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e065cd65ec00f8368f1fb3e237ce517ca0d19694fded47e8b3900325c085e044"},"source":{"id":"2510.18034","kind":"arxiv","version":3},"verdict":{"id":"c52c4db2-f984-44e3-963a-df5747e99981","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-18T05:43:04.751411Z","strongest_claim":"Applying SAVANT improves VLM's absolute recall by approximately 18.5% compared to prompting baselines, and leveraging the best proprietary model within the framework enables automatic labeling of around 10,000 images to fine-tune a 7B open-source model achieving 90.8% recall and 93.8% accuracy.","one_line_summary":"SAVANT boosts VLM recall for semantic anomaly detection in driving images by 18.5% via structured reasoning and enables fine-tuning a 7B open model to 90.8% recall and 93.8% accuracy.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The evaluation uses a 'balanced set of real-world driving scenarios' whose selection criteria and representativeness of long-tail anomalies are not specified, which is required to support that the reported recall gains are due to the structured reasoning pipeline rather than dataset construction.","pith_extraction_headline":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2510.18034/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":2,"snapshot_sha256":"2fceb3457930879b5343fa37b2737a3e105095b36484eb43a163ae2953f3523f"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"c52c4db2-f984-44e3-963a-df5747e99981"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-21T01:05:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QJvgd5niaLZb/s0uyrAtxcax7Gv7NO+4DuWmoXTVTtFHP7zuT+tJ91XDy2no4v9NjU8ERn8eMeQvGKXXeutiCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T05:34:10.170086Z"},"content_sha256":"11fda04308e8063f37fa2ef23bbd907da87d65ec906ed04885aa3a5c4299fc02","schema_version":"1.0","event_id":"sha256:11fda04308e8063f37fa2ef23bbd907da87d65ec906ed04885aa3a5c4299fc02"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/bundle.json","state_url":"https://pith.science/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/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":"2026-05-26T05:34:10Z","links":{"resolver":"https://pith.science/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y","bundle":"https://pith.science/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/bundle.json","state":"https://pith.science/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PEBP4TFO5XRL37FWVRS5A2Q35Y/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:PEBP4TFO5XRL37FWVRS5A2Q35Y","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":"593f05e84dc7ad1f8d210d0208c7e01fd913cc1d269b611aad76b7acca25ea21","cross_cats_sorted":["cs.AI","cs.RO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-10-20T19:14:29Z","title_canon_sha256":"aa1b55cc34fcf2b2d0ea87381c2ee6c38355f7435c3639f9b8b6cb07a037c177"},"schema_version":"1.0","source":{"id":"2510.18034","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.18034","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"arxiv_version","alias_value":"2510.18034v3","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.18034","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_12","alias_value":"PEBP4TFO5XRL","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_16","alias_value":"PEBP4TFO5XRL37FW","created_at":"2026-05-21T01:05:11Z"},{"alias_kind":"pith_short_8","alias_value":"PEBP4TFO","created_at":"2026-05-21T01:05:11Z"}],"graph_snapshots":[{"event_id":"sha256:11fda04308e8063f37fa2ef23bbd907da87d65ec906ed04885aa3a5c4299fc02","target":"graph","created_at":"2026-05-21T01:05:11Z","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":"Applying SAVANT improves VLM's absolute recall by approximately 18.5% compared to prompting baselines, and leveraging the best proprietary model within the framework enables automatic labeling of around 10,000 images to fine-tune a 7B open-source model achieving 90.8% recall and 93.8% accuracy."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The evaluation uses a 'balanced set of real-world driving scenarios' whose selection criteria and representativeness of long-tail anomalies are not specified, which is required to support that the reported recall gains are due to the structured reasoning pipeline rather than dataset construction."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"SAVANT boosts VLM recall for semantic anomaly detection in driving images by 18.5% via structured reasoning and enables fine-tuning a 7B open model to 90.8% recall and 93.8% accuracy."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall."}],"snapshot_sha256":"e065cd65ec00f8368f1fb3e237ce517ca0d19694fded47e8b3900325c085e044"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"2fceb3457930879b5343fa37b2737a3e105095b36484eb43a163ae2953f3523f"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2510.18034/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Autonomous driving systems remain critically vulnerable to the long-tail of rare, out-of-distribution semantic anomalies. While VLMs have emerged as promising tools for perception, their application in anomaly detection remains largely restricted to prompting proprietary models - limiting reliability, reproducibility, and deployment feasibility. To address this gap, we introduce SAVANT (Semantic Anomaly Verification/Analysis Toolkit), a novel model-agnostic reasoning framework that reformulates anomaly detection as a layered semantic consistency verification. By applying SAVANT's two-phase pip","authors_text":"David Pop, Johannes Betz, Mattia Piccinini, Roberto Brusnicki, Yuan Gao","cross_cats":["cs.AI","cs.RO"],"headline":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-10-20T19:14:29Z","title":"Can VLMs Unlock Semantic Anomaly Detection? A Framework for Structured Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.18034","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-18T05:43:04.751411Z","id":"c52c4db2-f984-44e3-963a-df5747e99981","model_set":{"reader":"grok-4.3"},"one_line_summary":"SAVANT boosts VLM recall for semantic anomaly detection in driving images by 18.5% via structured reasoning and enables fine-tuning a 7B open model to 90.8% recall and 93.8% accuracy.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Structured reasoning framework lets VLMs detect semantic anomalies in driving scenes with 18.5 percent higher recall.","strongest_claim":"Applying SAVANT improves VLM's absolute recall by approximately 18.5% compared to prompting baselines, and leveraging the best proprietary model within the framework enables automatic labeling of around 10,000 images to fine-tune a 7B open-source model achieving 90.8% recall and 93.8% accuracy.","weakest_assumption":"The evaluation uses a 'balanced set of real-world driving scenarios' whose selection criteria and representativeness of long-tail anomalies are not specified, which is required to support that the reported recall gains are due to the structured reasoning pipeline rather than dataset construction."}},"verdict_id":"c52c4db2-f984-44e3-963a-df5747e99981"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:128c6f264134f31586bb21ce4b4c2b9c1991f2b74c894bb72fc6549d6665d2c7","target":"record","created_at":"2026-05-21T01:05:11Z","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":"593f05e84dc7ad1f8d210d0208c7e01fd913cc1d269b611aad76b7acca25ea21","cross_cats_sorted":["cs.AI","cs.RO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-10-20T19:14:29Z","title_canon_sha256":"aa1b55cc34fcf2b2d0ea87381c2ee6c38355f7435c3639f9b8b6cb07a037c177"},"schema_version":"1.0","source":{"id":"2510.18034","kind":"arxiv","version":3}},"canonical_sha256":"7902fe4caeede2bdfcb6ac65d06a1bee3be3190dc6b428d4b303493b606df8d6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7902fe4caeede2bdfcb6ac65d06a1bee3be3190dc6b428d4b303493b606df8d6","first_computed_at":"2026-05-21T01:05:11.175417Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T01:05:11.175417Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fKIjdb15PYV3eV5EhYl4CDt6FOTldCogDrFwF53tzZZ9Dd3/zV4IeqHXS0tpUtwD1pshCDiifvbH9Dkk8eZZAw==","signature_status":"signed_v1","signed_at":"2026-05-21T01:05:11.176306Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.18034","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:128c6f264134f31586bb21ce4b4c2b9c1991f2b74c894bb72fc6549d6665d2c7","sha256:11fda04308e8063f37fa2ef23bbd907da87d65ec906ed04885aa3a5c4299fc02"],"state_sha256":"606285241c37de40bfcda748463591a5201087171f7acc99827966288f49fd66"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CmGzWsJ2PiSCQKoxdIv2ccN6qh++5Zy9wCA+38n2i7HLAP5//v/+ZESVsOb/BubXioBt8spVFqe2hrSLaYxECQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T05:34:10.174045Z","bundle_sha256":"8d63c258a38f1a80460f593b8093e33e452505b78daa459486e216f65cce1f90"}}