{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:45Z5HPR5RE2NNOKORQVTKGLC47","short_pith_number":"pith:45Z5HPR5","canonical_record":{"source":{"id":"2608.08029","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T09:34:22Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"442bb060ec0dd22257e6ae0bbc7d86fe262b93055401d90dae01a107ed57156b","abstract_canon_sha256":"760dadd1c5f6f12b9492a66cc3093e0521021576692cf4ebe8adb8339e1c746e"},"schema_version":"1.0"},"canonical_sha256":"e773d3be3d8934d6b94e8c2b351962e7cb717875b2a1adcc153cebc36c228ff8","source":{"kind":"arxiv","id":"2608.08029","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.08029","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"arxiv_version","alias_value":"2608.08029v1","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08029","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_12","alias_value":"45Z5HPR5RE2N","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_16","alias_value":"45Z5HPR5RE2NNOKO","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_8","alias_value":"45Z5HPR5","created_at":"2026-08-11T01:21:26Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:45Z5HPR5RE2NNOKORQVTKGLC47","target":"record","payload":{"canonical_record":{"source":{"id":"2608.08029","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T09:34:22Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"442bb060ec0dd22257e6ae0bbc7d86fe262b93055401d90dae01a107ed57156b","abstract_canon_sha256":"760dadd1c5f6f12b9492a66cc3093e0521021576692cf4ebe8adb8339e1c746e"},"schema_version":"1.0"},"canonical_sha256":"e773d3be3d8934d6b94e8c2b351962e7cb717875b2a1adcc153cebc36c228ff8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:21:26.898278Z","signature_b64":"ObC9qgwW2b3qAj4BnTsbZfSrE60QaS4CtrW7sPJLB9P8pDrn+X/4031mm2bwcvgVaRvk8cYHIm21JPHi9ucoAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e773d3be3d8934d6b94e8c2b351962e7cb717875b2a1adcc153cebc36c228ff8","last_reissued_at":"2026-08-11T01:21:26.895873Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:21:26.895873Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.08029","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-11T01:21:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"I6k4eNy85ahqmQDla+CL3pYXQWP8+HLJBl6g0C8+g50H0BGOIW2BpEOcUFMaS/hTFVGs/vpwfQD9PBYbJ5EKDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T01:07:36.060493Z"},"content_sha256":"6fba76375630e7a39104d9b5ffebc66e0b79847cb02b19b54382e971b72bccfe","schema_version":"1.0","event_id":"sha256:6fba76375630e7a39104d9b5ffebc66e0b79847cb02b19b54382e971b72bccfe"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:45Z5HPR5RE2NNOKORQVTKGLC47","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Do All LLMs Know When They're Being Harmful? A Reproducibility Study of Latent-Space Safety Probes Across Model Families","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Alizishaan Khatri, Dun Li Chan","submitted_at":"2026-08-08T09:34:22Z","abstract_excerpt":"Khatri et al. (2026) [DOI: 10.1109/DSN-W70714.2026.00027] show that lightweight MLP probes on final-layer activations of a single 8B model (LLaMA-3.1-8B) detect harmful prompts at F1 competitive with guard models 1000x larger, using one probe per benchmark. We reproduce this pipeline end-to-end and extend it along two axes the original study leaves open. First, we test whether the result generalizes across other model architecture and scale by training identical probes on activations from models like Gemma-4-E4B, Mistral-7B-v0.3, and Qwen2-7B, using the three benchmarks (WildJailbreak, BeaverT"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08029","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/2608.08029/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":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-11T01:21:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"etVYk7R4WCufs/jgW8FQRiEJqzTIAMwe24ZTndszGZ3+gpn5JKNVZ90+/LUP+gejaVo+uS7Fh+t7HOkGdxpwCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T01:07:36.061039Z"},"content_sha256":"e7b556676bca7425449920ef573efab80a72d003ffb39b91fad3f551f8741cec","schema_version":"1.0","event_id":"sha256:e7b556676bca7425449920ef573efab80a72d003ffb39b91fad3f551f8741cec"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/45Z5HPR5RE2NNOKORQVTKGLC47/bundle.json","state_url":"https://pith.science/pith/45Z5HPR5RE2NNOKORQVTKGLC47/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/45Z5HPR5RE2NNOKORQVTKGLC47/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":"2026-08-20T01:07:36Z","links":{"resolver":"https://pith.science/pith/45Z5HPR5RE2NNOKORQVTKGLC47","bundle":"https://pith.science/pith/45Z5HPR5RE2NNOKORQVTKGLC47/bundle.json","state":"https://pith.science/pith/45Z5HPR5RE2NNOKORQVTKGLC47/state.json","well_known_bundle":"https://pith.science/.well-known/pith/45Z5HPR5RE2NNOKORQVTKGLC47/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:45Z5HPR5RE2NNOKORQVTKGLC47","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":"760dadd1c5f6f12b9492a66cc3093e0521021576692cf4ebe8adb8339e1c746e","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T09:34:22Z","title_canon_sha256":"442bb060ec0dd22257e6ae0bbc7d86fe262b93055401d90dae01a107ed57156b"},"schema_version":"1.0","source":{"id":"2608.08029","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.08029","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"arxiv_version","alias_value":"2608.08029v1","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08029","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_12","alias_value":"45Z5HPR5RE2N","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_16","alias_value":"45Z5HPR5RE2NNOKO","created_at":"2026-08-11T01:21:26Z"},{"alias_kind":"pith_short_8","alias_value":"45Z5HPR5","created_at":"2026-08-11T01:21:26Z"}],"graph_snapshots":[{"event_id":"sha256:e7b556676bca7425449920ef573efab80a72d003ffb39b91fad3f551f8741cec","target":"graph","created_at":"2026-08-11T01:21:26Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2608.08029/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Khatri et al. (2026) [DOI: 10.1109/DSN-W70714.2026.00027] show that lightweight MLP probes on final-layer activations of a single 8B model (LLaMA-3.1-8B) detect harmful prompts at F1 competitive with guard models 1000x larger, using one probe per benchmark. We reproduce this pipeline end-to-end and extend it along two axes the original study leaves open. First, we test whether the result generalizes across other model architecture and scale by training identical probes on activations from models like Gemma-4-E4B, Mistral-7B-v0.3, and Qwen2-7B, using the three benchmarks (WildJailbreak, BeaverT","authors_text":"Alizishaan Khatri, Dun Li Chan","cross_cats":["cs.AI","cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T09:34:22Z","title":"Do All LLMs Know When They're Being Harmful? A Reproducibility Study of Latent-Space Safety Probes Across Model Families"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08029","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:6fba76375630e7a39104d9b5ffebc66e0b79847cb02b19b54382e971b72bccfe","target":"record","created_at":"2026-08-11T01:21:26Z","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":"760dadd1c5f6f12b9492a66cc3093e0521021576692cf4ebe8adb8339e1c746e","cross_cats_sorted":["cs.AI","cs.CR"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-08-08T09:34:22Z","title_canon_sha256":"442bb060ec0dd22257e6ae0bbc7d86fe262b93055401d90dae01a107ed57156b"},"schema_version":"1.0","source":{"id":"2608.08029","kind":"arxiv","version":1}},"canonical_sha256":"e773d3be3d8934d6b94e8c2b351962e7cb717875b2a1adcc153cebc36c228ff8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e773d3be3d8934d6b94e8c2b351962e7cb717875b2a1adcc153cebc36c228ff8","first_computed_at":"2026-08-11T01:21:26.895873Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-11T01:21:26.895873Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ObC9qgwW2b3qAj4BnTsbZfSrE60QaS4CtrW7sPJLB9P8pDrn+X/4031mm2bwcvgVaRvk8cYHIm21JPHi9ucoAA==","signature_status":"signed_v1","signed_at":"2026-08-11T01:21:26.898278Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.08029","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6fba76375630e7a39104d9b5ffebc66e0b79847cb02b19b54382e971b72bccfe","sha256:e7b556676bca7425449920ef573efab80a72d003ffb39b91fad3f551f8741cec"],"state_sha256":"aa8da152324a978fd772173b33efca2c70e10f07c70d5eee6a3d9a3e7bca99a4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3Xb8ZER+3i4o1yWHcDpnfG33EZEXsK/LRD0ofKEN9XpDTu6F1/KK5EoF13LrWTK00ifkWNWhhHe7DueI00aMCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T01:07:36.066825Z","bundle_sha256":"53cafae82343f9fc0df9db7f55f42b9b94a4960c28b4df3e95219049c1067530"}}