{"_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_a1149f4f114eecd727e873765b64065c2e356fa1b9a0eeeeb9726d410007a0aa","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_a1149f4f114eecd727e873765b64065c2e356fa1b9a0eeeeb9726d410007a0aa","axiom_type":"AX.OBS","body":{"axiom_subtype":"news.vendor_press.v1","category":"news","fingerprint":"f1bef5c9744ea5f0fbde91b4e7852502589ef9e6bb8c56bce9901fcd2c82eb42","published":"Wed, 20 May 2026 00:00:00 -0400","receipt_hash":"f1bef5c9744ea5f0fbde91b4e7852502589ef9e6bb8c56bce9901fcd2c82eb42","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":"f1bef5c9744ea5f0fbde91b4e7852502589ef9e6bb8c56bce9901fcd2c82eb42","observed_at":"2026-05-20T04:43:44.562035Z","parent_run_hash":"5f904c2c2fecb6b44f2adce8bdc9de914b9a39f7c4fdd7bf086a6c2361a30f8c","published":"Wed, 20 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:2602.14778v3 Announce Type: replace-cross \nAbstract: Hallucinations -- plausible but factually incorrect responses -- pose a major challenge to the reliability of Large Language Models (LLMs), especially in multi-step or agentic settings.\n  Existing work largely frames hallucinations as a consequence of missing knowledge; we show instead that, even when the relevant factual knowledge is present, models still produce hallucinated answers, pointing to retrieval instability rather than knowledge gaps.\n  Building on this observation, we introduce APORIA (Aggregate Prompt-wise Observation Retrieving Instability via Asymmetry -- the state of puzzlement-in-contradiction that hallucinations embody), a geometric framework that studies repeated responses to the same prompt in sentence-embedding space. Our central hypothesis is that genuine responses cluster more tightly than hallucinated ones; we empirically validate this and show that, after Fisher projection, the two response classes bec","title":"A Geometric Analysis of Small-sized Language Model Hallucinations","url":"https://arxiv.org/abs/2602.14778","vendor":"arxiv_cs_ai"},"summary":"arXiv:2602.14778v3 Announce Type: replace-cross \nAbstract: Hallucinations -- plausible but factually incorrect responses -- pose a major challenge to the reliability of Large Language Models (LLMs), especially in multi-step or agentic settings.\n  Existing work largely frames hallucinations as a consequence of missing knowledge; we show instead that, even when the relevant factual knowledge is present, models still produce hallucinated answers, pointing to retrieval instability rather than knowledge gaps.\n  Building on this observation, we introduce APORIA (Aggregate Prompt-wise Observation Retrieving Instability via Asymmetry -- the state of puzzlement-in-contradiction that hallucinations embody), a geometric framework that studies repeated responses to the same prompt in sentence-embedding space. Our central hypothesis is that genuine responses cluster more tightly than hallucinated ones; we empirically validate this and show that, after Fisher projection, the two response classes bec","title":"A Geometric Analysis of Small-sized Language Model Hallucinations","vendor":"arxiv_cs_ai"},"confidence":{"method":"deterministic"},"decision":"POSITIVE","issued_at":"2026-05-20T04:43:44Z","notes":"Spider vendor_press (news) news.vendor_press.v1","object":{"captured_by":"crovia.spider.vendor_press","primary_source_url":"https://arxiv.org/abs/2602.14778"},"predecessors":[],"schema":"crovia.axiom.v1","signature":"ed25519:319846f53b7ef920e6f85eae1c7a528aa743f58b9ce594ba7ea53e2c627b634eee98b023f65f6237da35846f0601846d253281c994c15f42ffe7a077635b310a","signer":"crovia.substrate","subject":{"observed_at":"2026-05-20T04:43:44Z","source_collector":"spider:vendor_press","target_id":"https://arxiv.org/abs/2602.14778"},"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":"9054e1ce98c8e2c74d847ff4d13ca48f0911775d280759aebf80212aba85c0b5","leaf_index":145293,"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":"15f0cbf4797cf7ab3deb68ad7a19c58df40e96640670bd1358085c39375b0a52","side":"left"},{"sibling":"fbe49736b14df07062ba81087eb1d33c3e4c26b1be72a48815e5a06854fdf8fb","side":"right"},{"sibling":"369fe1e6c4d91bf7a1bc443a5337c64c458a601160c7954ca0df942cf631374a","side":"left"},{"sibling":"3843aae7674dec39b1c581bc4b4aa0271812fceccf34cb1c6d1439214ab8543d","side":"left"},{"sibling":"5f03ce43876e6319740b03425a5aeaeb109f248a221d0d33fd4eee59673f01e9","side":"right"},{"sibling":"01ed1e09ca55b8947e9e2a49f489627f1a042f0c03eac8fb97fed38389096fa5","side":"right"},{"sibling":"cefd93338369caa43fee7488c87e004962f685c84a7eb2dabfe358e69b38d6f1","side":"right"},{"sibling":"cc6b64c1eb047b460e9fe9beb452cf69fe3eb490dcf6d93d8193dda021cdf8ef","side":"left"},{"sibling":"f073aa7be27ee7e1d9eb0de5f129f7e9bae0c584fb959378c395b65831dc1d1d","side":"left"},{"sibling":"1526885f19d1fadf6955cf519dbc4e62d593a4bba99d741e8c301740a7068233","side":"left"},{"sibling":"e3a7d5c07f161682d61bd453ffc02ecdf87cfeda70f986d6650017f9d2d6b265","side":"left"},{"sibling":"edbc49f08e5b92291934c05c9e6efd270a6b0698d8d2fa474006366027dfe098","side":"right"},{"sibling":"3e4df6e7457cecbf36f350375e72dcab336a3984422e4c406ef809e4e2944e96","side":"left"},{"sibling":"8f4c0fbe56b6c010fbb8c782ebcd478079bb3f991d8704e2534209a075d9163c","side":"left"},{"sibling":"be08fedc4e72a6fb56606f66812fae7317e09690b9acb18385f4ab117a981238","side":"right"},{"sibling":"0534329a7475dc9df51998c83c16892126679dade0fa34182f21e869599386c7","side":"right"},{"sibling":"2d24720928ead0e7670650eb55f558c4f20e4c18df376f47ba72cfa8cf0ed344","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":147301,"merkle_root":"08903d7159c3b38eeeeafc09eab15139ea417f1d94f02f1fbc87296b37db840a","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","run_id":"hourly_json_retrofit_20260521T183701Z","schema":"crovia.seal.v1","seal_family_version":"crovia-seal-family/1","seal_kind":"substrate_batch","sealed_at":"2026-05-21T18:37:33Z","sig_algorithm":"ed25519","signature":"905f2924632dfa2970c8690285f5b5d4a1d891d0e0ef1cbc404ebec2fd937215ac768e16f0a9f28b18977a55ae0bfd226db6833ae7ef588729054117d2da7303","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_a1149f4f114eecd727e873765b64065c2e356fa1b9a0eeeeb9726d410007a0aa"}}