{"_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_b2b8317515a86a7b9265aec0d31b22cfebc0b04359e5c99b523f724f1894cad3","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_b2b8317515a86a7b9265aec0d31b22cfebc0b04359e5c99b523f724f1894cad3","axiom_type":"AX.OBS","body":{"axiom_subtype":"news.vendor_press.v1","category":"news","fingerprint":"334dd59d8fad565bbb7c718d0b8b60b6e1988f32b4ad05f4a3cf46e0dad2a4f6","published":"Mon, 15 Jun 2026 00:00:00 -0400","receipt_hash":"334dd59d8fad565bbb7c718d0b8b60b6e1988f32b4ad05f4a3cf46e0dad2a4f6","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":"334dd59d8fad565bbb7c718d0b8b60b6e1988f32b4ad05f4a3cf46e0dad2a4f6","observed_at":"2026-06-15T04:43:09.998079Z","parent_run_hash":"ded7a5fa7968821af82d6d8d24b2c1f7e7d776433180609016edbdee95e78c1a","published":"Mon, 15 Jun 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:2606.13685v1 Announce Type: cross \nAbstract: LLM-as-a-Judge is now widely used to rank model outputs, train reward models, and populate public leaderboards, but its run-to-run reliability remains under-characterized. We study repeated identical evaluations on 29 tasks spanning 10 categories using two OpenAI judge models (GPT-4o-mini and GPT-4.1-mini), with 50 pairwise trials and 50 pointwise trials per question, supplemented by temperature and prompt-sensitivity ablations. Across judges, pairwise preferences flip on average 13.6% of the time, with 28% of questions exceeding a 20% flip rate and one question reaching 56%. GPT-4o-mini also exhibits a significant first-position bias (72% A-majority, p = 0.024). At the same time, mean pointwise score gaps are small (0.19--0.36 on a 10-point scale) and not statistically significant in aggregate, producing a pairwise--pointwise gap: judges frequently choose a winner even when their own scalar scores provide little evidence of a meaningf","title":"The Coin Flip Judge? Reliability and Bias in LLM-as-a-Judge Evaluation","url":"https://arxiv.org/abs/2606.13685","vendor":"arxiv_cs_ai"},"summary":"arXiv:2606.13685v1 Announce Type: cross \nAbstract: LLM-as-a-Judge is now widely used to rank model outputs, train reward models, and populate public leaderboards, but its run-to-run reliability remains under-characterized. We study repeated identical evaluations on 29 tasks spanning 10 categories using two OpenAI judge models (GPT-4o-mini and GPT-4.1-mini), with 50 pairwise trials and 50 pointwise trials per question, supplemented by temperature and prompt-sensitivity ablations. Across judges, pairwise preferences flip on average 13.6% of the time, with 28% of questions exceeding a 20% flip rate and one question reaching 56%. GPT-4o-mini also exhibits a significant first-position bias (72% A-majority, p = 0.024). At the same time, mean pointwise score gaps are small (0.19--0.36 on a 10-point scale) and not statistically significant in aggregate, producing a pairwise--pointwise gap: judges frequently choose a winner even when their own scalar scores provide little evidence of a meaningf","title":"The Coin Flip Judge? Reliability and Bias in LLM-as-a-Judge Evaluation","vendor":"arxiv_cs_ai"},"confidence":{"method":"deterministic"},"decision":"POSITIVE","issued_at":"2026-06-15T04:43:09Z","notes":"Spider vendor_press (news) news.vendor_press.v1","object":{"captured_by":"crovia.spider.vendor_press","primary_source_url":"https://arxiv.org/abs/2606.13685"},"predecessors":[],"schema":"crovia.axiom.v1","signature":"ed25519:4d3f9cda786b751aeffc72484acbf0ef7ff5ce677f6e0124e35e70129b7cbbf826e08dad6b62172145369c5187985e516d190a111913238ebaa7e0aa04bb270f","signer":"crovia.substrate","subject":{"observed_at":"2026-06-15T04:43:09Z","source_collector":"spider:vendor_press","target_id":"https://arxiv.org/abs/2606.13685"},"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":"7a39e4ead2fa4e1ad3f34f9a3bda852df52ea33c34b24468a5280ab25eb96d6d","leaf_index":229967,"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":"b3797795fa2febf47b8b717f3b3821df8b96b669df8835e51a528733af731314","side":"left"},{"sibling":"4285e1e66dfae27042ea677842118c8790f0dcaec2fee1929aa2887825ae3815","side":"left"},{"sibling":"fc39a72a5a1f754bd58a69833e00a9659a3961f8e21250ec3d5d5a236ef3ae91","side":"left"},{"sibling":"7a3924f22fb8cafad2ee97bf7cbbc8a87ae838940d515a7387c7fa1c239bbc64","side":"left"},{"sibling":"20a16b9950c14e7c13f1fe96b75ed0ebe69644ed4b1d06f1f3a2b4f5c64e86dd","side":"right"},{"sibling":"861237052056050d2c3d73bb49e6445a4c553842a8208d157d0049c55f967800","side":"right"},{"sibling":"117727e77823a42d7a3da4dfbc182b574e2d826f75958c69340ae784488a0cfe","side":"left"},{"sibling":"bc1002bb7e3da047b8a7c8f3990db61fb56edf499868605dc0c31aa0dee387b7","side":"right"},{"sibling":"14c50c43949e1ad41f149ffea691627d3f715c5861766c693b9fbac9d03b0d90","side":"right"},{"sibling":"74897e850164dddc689c3c65b33f9bae0268ab0bf429867a4e193d9b9b685040","side":"left"},{"sibling":"bde25d7e94e64717e426a97f6fcb4907e92b5c61fc89d92d7e0947a2249c3f6b","side":"right"},{"sibling":"d10d772a4984cae00e65ab24af21d1d260e475bbaae3f17871859eb705bd3999","side":"right"},{"sibling":"e79159853f2f35ddae8e3247d515e433c534277b287d65bbd77ae989aa4992fa","side":"right"},{"sibling":"3054319f1840cce0eaaf0bc4b1ae38e5bf8b6210927d924a750775cc7d77cca6","side":"right"},{"sibling":"0fd8b5059f279c4a4a6688de2472fdbc25543fee183df19b21dacca880354cff","side":"right"},{"sibling":"a04392fb9f2a3a620840e3b3fecd93d12e6d224e481c162389d8ae64e7194599","side":"left"},{"sibling":"c300cf0154c136afc09b1702a0be98f4ba5b6dc5cf57e8cc714ec1eaf4196eff","side":"left"},{"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":232015,"merkle_root":"62bfb7809bb55667ad7eeebdb48267b9b2c1ee89bb808ea1e6d3a814a15aa402","public_key_hex":"cf742e26f75669dc673cb5c0786a1ae23ae8ca19c347317192ce40c28a7ff25c","run_id":"hourly_json_retrofit_20260617T133701Z","schema":"crovia.seal.v1","seal_family_version":"crovia-seal-family/1","seal_kind":"substrate_batch","sealed_at":"2026-06-17T13:39:59Z","sig_algorithm":"ed25519","signature":"930a3563c643cc7518d048b12a1f5392a96a5edce49533ba0a56f11b6cc69319bb1c800aacc46b52a6941c4d05dae255a45cac757d6093c971b96852c24f140e","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_b2b8317515a86a7b9265aec0d31b22cfebc0b04359e5c99b523f724f1894cad3"}}