{"_canonicalization":{"envelope_id":"axm_ + sha256(envelope minus {signature, axiom_id, anchors})","envelope_signature":"ed25519(envelope minus {signature, axiom_id})","json":"sort_keys=True, separators=(',',':'), ensure_ascii=False, allow_nan=False, utf-8","leaf_hash":"sha256(0x00 || canonical_json(envelope_full))","seal_signature":"ed25519(seal minus {signature, sig_algorithm})"},"axiom_id":"axm_a4e8277f1690386c032f4437d55a72d3fc98d978504337551c13695f7f78d7e2","bitcoin_anchor":{"bitcoin_attestations":[],"calendar_attestations":[],"ots_url":"","stamped_at":"","status":"pending_next_stamp"},"envelope":{"anchors":[{"chain":"crovia.axiom_graph","height":0,"merkle_proof":"spider_vendor_press_v1","root_at_anchor":"spider_vendor_press_v1"}],"axiom_id":"axm_a4e8277f1690386c032f4437d55a72d3fc98d978504337551c13695f7f78d7e2","axiom_type":"AX.OBS","body":{"axiom_subtype":"news.vendor_press.v1","category":"news","fingerprint":"fc52f88e970302f0ad43c0c1e41fe65f337ed78e410278e23dad8f4851613f9a","published":"Thu, 28 May 2026 00:00:00 -0400","receipt_hash":"fc52f88e970302f0ad43c0c1e41fe65f337ed78e410278e23dad8f4851613f9a","schema":"spider.news.vendor_press.v1","spider":"vendor_press","spider_record":{"axiom_subtype":"news.vendor_press.v1","category":"news","decision_hint":"POSITIVE","envelope_target":"AX.OBS","fingerprint":"fc52f88e970302f0ad43c0c1e41fe65f337ed78e410278e23dad8f4851613f9a","observed_at":"2026-05-28T04:43:38.862500Z","parent_run_hash":"58f8b4a134069e0a15ea3949252489597eb86dd27c9ca3fb15c6fb838ce49ef3","published":"Thu, 28 May 2026 00:00:00 -0400","runtime_version":"0.1.0","schema":"spider.news.vendor_press.v1","source_status":200,"source_url":"https://export.arxiv.org/rss/cs.AI","spider":"vendor_press","summary_excerpt":"arXiv:2605.27958v1 Announce Type: cross \nAbstract: Linear probes trained on LLM activations are increasingly proposed as deception-detection metrics, yet report AUROC exceeding 0.96 on clean benchmarks while collapsing under distributional shift. This paper systematically pressure-tests probe-based metrics across the Gemma 3 model family (1B-27B parameters), diagnosing why they fail rather than merely documenting that they fail. We test four hypotheses about deception encoding: (1) single linear direction, (2) multi-dimensional subspace, (3) convex conic hull, (4) entropy proxy. Our design includes cross-domain transfer matrices, multi-dimensional probe analysis with permutation null baselines, entropy-residualization tests, and distractor evaluations across 8 stylistic shifts. We find that: (a) probes achieve near-perfect AUROC (>=0.998) on clean data but collapse under stylistic shifts; style-augmented probes recover near-perfect detection (mean AUROC 0.979-0.983) on unseen styles; (","title":"Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations","url":"https://arxiv.org/abs/2605.27958","vendor":"arxiv_cs_ai"},"summary":"arXiv:2605.27958v1 Announce Type: cross \nAbstract: Linear probes trained on LLM activations are increasingly proposed as deception-detection metrics, yet report AUROC exceeding 0.96 on clean benchmarks while collapsing under distributional shift. This paper systematically pressure-tests probe-based metrics across the Gemma 3 model family (1B-27B parameters), diagnosing why they fail rather than merely documenting that they fail. We test four hypotheses about deception encoding: (1) single linear direction, (2) multi-dimensional subspace, (3) convex conic hull, (4) entropy proxy. Our design includes cross-domain transfer matrices, multi-dimensional probe analysis with permutation null baselines, entropy-residualization tests, and distractor evaluations across 8 stylistic shifts. We find that: (a) probes achieve near-perfect AUROC (>=0.998) on clean data but collapse under stylistic shifts; style-augmented probes recover near-perfect detection (mean AUROC 0.979-0.983) on unseen styles; (","title":"Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations","vendor":"arxiv_cs_ai"},"confidence":{"method":"deterministic"},"decision":"POSITIVE","issued_at":"2026-05-28T04:43:38Z","notes":"Spider vendor_press (news) news.vendor_press.v1","object":{"captured_by":"crovia.spider.vendor_press","primary_source_url":"https://arxiv.org/abs/2605.27958"},"predecessors":[],"schema":"crovia.axiom.v1","signature":"ed25519:fa16ce814c94cf070ae26200c7a9d6ea95eb5706e623f39c856d6dc9096e3ee10df66bf56e19b352da29389b9e5363aae68c252189e601db4c0f732b4a790007","signer":"crovia.substrate","subject":{"observed_at":"2026-05-28T04:43:38Z","source_collector":"spider:vendor_press","target_id":"https://arxiv.org/abs/2605.27958"},"tsa":{"authority":"crovia.substrate.bootstrap","rfc3161_token":"{\"kind\":\"crovia.bootstrap.tsa\",\"source_jsonl\":\"/opt/crovia/spider/data/news/vendor_press_v1.jsonl\",\"source_seal_merkle_root\":\"spider_vendor_press_v1\",\"upgrade_path\":\"Sessione H \\u2014 OpenTimestamps weekly anchor\"}"},"zk_mode":"clear","zk_proof":null},"ledger":{"leaf_hash":"faa3e5001caf95e53ab0706b2bfe2a90008d5e68d29b99129930d2976f220def","leaf_index":155840,"ledger_path":"/opt/crovia/substrate/axiom_ledger.jsonl"},"merkle_proof":{"hash_alg":"sha256","leaf_prefix":"0x00","node_prefix":"0x01","odd_leaf_rule":"duplicate_last","path":[{"sibling":"235d97457d978b5a2cff16b87f9540306453034cbff5729088e6024a8d47987b","side":"right"},{"sibling":"0c6d915e71b320fec111aa9ad3d5461c0d0b24f1c15a9c3bcef31f2fbef66a62","side":"right"},{"sibling":"7c6e0af114303b0eff6366707f95e4f867b7420dfc592ad332a8ec5223e63b15","side":"right"},{"sibling":"9ca138c42dbda55c2f57780f6467421febf19683db71e97728f75b80f122b31d","side":"right"},{"sibling":"47e8942202d3f064e04a8ef6adfe62e54351a41d826835564c748ce322eaf3bf","side":"right"},{"sibling":"025a0e48e2e601eb8df7d6306168114925eb79031fed301eaa235f39fb737e75","side":"right"},{"sibling":"23545a0a9721502eebf2f3a4b30acec3fc420534240d53ac993fddfb75131c5e","side":"left"},{"sibling":"753018fa645affe1bfe31d1c496292a6e72167116d741961e921b85a693e4153","side":"left"},{"sibling":"0adb35dd3ad9bfc4a781c02606acf0a3ce1e28e1a88066209cc1a35814e790ad","side":"right"},{"sibling":"f292d3278e493ec60902181b8c0bd5c89c0a1168ef928222fe6161287982f7f0","side":"right"},{"sibling":"2209295faf1a5bf51c97c6fd5a839a8181a4ea44f490420381530f35df7d9b2f","side":"right"},{"sibling":"5784576a15214ea9fc3569e6e1cff1ef443c0b1fc0d036028489088af089de27","side":"right"},{"sibling":"311772ec218efcb2da5a337f9e9f042fe1cc0028643adb0a354787e4ea7911b7","side":"right"},{"sibling":"66331bac84ca0f8983eb09fac7eaf95af234f1b82680b793eabff4ee25caac40","side":"left"},{"sibling":"35ca36cee447f0ef7064a25d55f59357c66901e31427729c4c1d8b14aa8adb6c","side":"left"},{"sibling":"7b6f0bea4291a5e63574dfca9aa0f9756f450c9478c3d07474d39a7ababb51f9","side":"right"},{"sibling":"1d39fe14b21e2ebbfb87e882423b24ee9469eae1e4c77af5b799ac4db9537467","side":"right"},{"sibling":"d841ad93efda0869e5eb97678f348f03f5caab4353e05ff4bf18f47fb945b822","side":"left"}]},"schema":"crovia.axiom_proof.v1","seal":{"first_collector_run_id":"","first_receipt_hash":"","jsonl_path":"/opt/crovia/substrate/axiom_ledger.jsonl","key_id":"430895f101d38164","last_collector_run_id":"","last_receipt_hash":"","leaf_count":156177,"merkle_root":"c5705a0243d16afd8b1ebfd731b7aa304079c442c2a7906493c5bbed374c69ec","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","run_id":"hourly_json_retrofit_20260528T053701Z","schema":"crovia.seal.v1","seal_family_version":"crovia-seal-family/1","seal_kind":"substrate_batch","sealed_at":"2026-05-28T05:37:36Z","sig_algorithm":"ed25519","signature":"f087e13febc8bb6a2e0812610de64cebc65be92915518d9c4b230799c3b161839b04c4eb1b02741f938f35545a76ab76555b04c782bdc2f9a44852d171d65909","signer_version":"1.1.0"},"trust_root":{"key_id":"430895f101d38164","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","signature_algorithm":"ed25519","url":"/registry/canon/TRUST_ROOT.md"},"verifier":{"spec":"/registry/canon/AXIOM_RECEIPT_v1.md","url":"/v/axm_a4e8277f1690386c032f4437d55a72d3fc98d978504337551c13695f7f78d7e2"}}