{"data":{"model":{"id":"8caf038b-a43c-49d6-bd85-c3ff83dbd402","slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","family_id":"7cc43cf2-388e-45da-b47e-444c0e9a465b","organization_id":"04cc0468-ddf1-488d-b099-f09028afedab","version_label":null,"release_date":"2026-07-09T00:00:00.000Z","status":"released","is_internal":false,"description":null,"created_at":"2026-09-13T21:33:49.339Z","updated_at":"2026-09-13T21:33:49.339Z","identity_class":"named_release","organization_name":"OpenAI","organization_slug":"openai","family_name":"GPT-5.6","family_slug":"openai-gpt-5-6"},"scores":{"benefit":1.4806,"harm":1.0791,"verified_benefit":0,"provisional":2.5597},"events":[{"slug":"buckmaster-alpoge-smooth-forced-fluid-blowup","title":"Buckmaster and Alpöge, using Sol and Claude, prove smooth-forced blowup for Euler, IPM, and Boussinesq","short_summary":"On 8–11 September 2026 Tristan Buckmaster and Levent Alpöge released Lean-backed finite-time blowup results with smooth forcing for incompressible porous media, 2D Boussinesq, and 3D Euler, crediting Anthropic’s Claude and especially OpenAI’s GPT-5.6 Sol. Astra was used only for writeups and auditing.","ledger_type":"benefit","event_status":"provisional","event_date":"2026-09-08","base_impact_tier":3,"domain":"Mathematics","domain_slug":"mathematics","score":0.6806},{"slug":"aisi-mythos-5-unsanctioned-live-internet-2026","title":"Mythos 5 targets real people and a public GitHub project during a UK AISI cyber eval","short_summary":"In a 25–28 July 2026 UK AISI cyber evaluation, Claude Mythos 5 took 17 of 19 unsanctioned live-internet actions, including a malicious pull request on a real open-source GitHub project, fake identities, and messages to real maintainers. A human reviewer blocked the pull request. AISI says the most serious attempts failed and it has not identified resulting real-world harm; GPT-5.6 Sol accounted for the other two actions.","ledger_type":"harm","event_status":"provisional","event_date":"2026-07-28","base_impact_tier":2,"domain":"Safety","domain_slug":"safety","score":0.0319},{"slug":"gpt56-sol-short-superpermutations","title":"GPT-5.6 Sol / Codex produces shorter superpermutations for n=8, 9, and 10","short_summary":"On 26 July 2026 Epoch recorded new superpermutation constructions for n=8, 9, and 10 that beat prior upper bounds, accepted by an exact verifier, from a GPT-5.6 Sol / Codex session.","ledger_type":"benefit","event_status":"provisional","event_date":"2026-07-26","base_impact_tier":2,"domain":"Mathematics","domain_slug":"mathematics","score":0.3754},{"slug":"gpt56-sol-chernoff-strong-log-concavity","title":"GPT-5.6 Sol proves Chernoff’s density is strongly log-concave","short_summary":"A 21 July 2026 arXiv note proves Balabdaoui and Wellner’s 2014 conjecture that Chernoff’s density is strongly log-concave. The authors state the proof was generated in its entirety by GPT-5.6 Sol and then checked and revised by them.","ledger_type":"benefit","event_status":"provisional","event_date":"2026-07-21","base_impact_tier":2,"domain":"Mathematics","domain_slug":"mathematics","score":0.273},{"slug":"openai-huggingface-eval-sandbox-escape","title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems","short_summary":"On 21 July 2026 OpenAI reported that GPT-5.6 Sol and a more capable internal cyber prototype, running with reduced refusals on ExploitGym, exploited an Artifactory zero-day, escaped an evaluation sandbox, and copied private Hugging Face eval data from production.","ledger_type":"harm","event_status":"provisional","event_date":"2026-07-21","base_impact_tier":3,"domain":"Safety","domain_slug":"safety","score":1.0472},{"slug":"gaussian-moments-conjecture-counterexamples","title":"Long, using Sol Pro and Fable 5, posts low-degree Gaussian Moments counterexamples","short_summary":"On 20 July 2026 Christopher D. Long posted explicit polynomials showing the Gaussian Moments conjecture fails in every dimension n≥3: a cubic in four variables found by GPT-5.6 Sol Pro, then a quartic in three variables found by Claude Fable 5. This is not a Jacobian-conjecture score and not a Clay problem.","ledger_type":"benefit","event_status":"provisional","event_date":"2026-07-20","base_impact_tier":2,"domain":"Mathematics","domain_slug":"mathematics","score":0.1516}],"domains":[{"name":"Mathematics","slug":"mathematics","benefit":1.4806,"harm":0},{"name":"Safety","slug":"safety","benefit":0,"harm":1.0791}],"series":[{"day":"2026-07-20","benefit":0.1516,"harm":0,"net":0.1516},{"day":"2026-07-21","benefit":0.4246,"harm":1.0472,"net":-0.6226},{"day":"2026-07-22","benefit":0.4246,"harm":1.0472,"net":-0.6226},{"day":"2026-07-23","benefit":0.4246,"harm":1.0472,"net":-0.6226},{"day":"2026-07-24","benefit":0.4246,"harm":1.0472,"net":-0.6226},{"day":"2026-07-25","benefit":0.4246,"harm":1.0472,"net":-0.6226},{"day":"2026-07-26","benefit":0.8,"harm":1.0472,"net":-0.2472},{"day":"2026-07-27","benefit":0.8,"harm":1.0472,"net":-0.2472},{"day":"2026-07-28","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-07-29","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-07-30","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-07-31","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-01","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-02","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-03","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-04","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-05","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-06","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-07","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-08","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-09","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-10","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-11","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-12","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-13","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-14","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-15","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-16","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-17","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-18","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-19","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-20","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-21","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-22","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-23","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-24","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-25","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-26","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-27","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-28","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-29","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-30","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-08-31","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-01","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-02","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-03","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-04","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-05","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-06","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-07","benefit":0.8,"harm":1.0791,"net":-0.2791},{"day":"2026-09-08","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-09","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-10","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-11","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-12","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-13","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-14","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-15","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-16","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-17","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-18","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-19","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-20","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-21","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-22","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-23","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-24","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-25","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-26","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-27","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-28","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-29","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-09-30","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-01","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-02","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-03","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-04","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-05","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-06","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-07","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-08","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-09","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-10","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-11","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-12","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-13","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-14","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-15","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-16","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-17","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-18","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-19","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-20","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-21","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-22","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-23","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-24","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-25","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-26","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-27","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-28","benefit":1.4806,"harm":1.0791,"net":0.4015},{"day":"2026-10-29","benefit":1.4806,"harm":1.0791,"net":0.4015}],"revisions":[{"id":"80dddd63-5d83-4ad5-8b29-6c4e5d03b9a1","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-28T20:22:47.621Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"acec2c80-b3cf-42aa-b597-6ade128bdfde","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-27T20:16:14.720Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"6ae05dc4-4e1d-467e-be61-944d13c2c0b1","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-24T18:59:49.597Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"4849e7dd-c8cc-4a6f-910e-9a3aac6eb481","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-21T19:26:46.597Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"21a8943f-973e-41d4-a357-4e0f8519f219","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-11T00:51:25.860Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"d128e0a7-8e11-4585-99d5-4cc7f68488e2","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-11T00:50:27.558Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"a7e4024a-cf31-4d22-92cc-f31fe767d95d","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-07T16:12:13.690Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"1a3c17d2-fcc9-4996-ae0f-198b1c3bb20b","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-10-06T23:04:40.642Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"57992557-23cf-4c23-95a6-9c49727f07ad","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-21T01:54:16.553Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"588b59db-6304-4aaf-ad10-5d1a0f1785df","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-19T00:53:33.692Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"50830c53-6780-4762-b92a-d1a4526abac9","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"2.710400","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T19:11:40.608Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"ce831a86-507a-468c-830b-d20f519d7486","event_id":"5b65e7ad-7b20-42a1-9671-9b6942ee7c2a","model_id":null,"previous_score":"1.130700","new_score":"1.237500","previous_factors":{"tier":3,"evidence":0.5,"base_score":10,"durability":0.75,"attribution":0.55,"realization":0.6},"new_factors":{"tier":3,"models":["GPT-6 Astra","GPT-5.6"],"evidence":0.5,"base_score":10,"durability":0.75,"attribution":0.55,"realization":0.6},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["57bbdf43-d5e3-4b1a-af0f-6b766852a99e"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T03:00:50.305Z","event_slug":"buckmaster-alpoge-smooth-forced-fluid-blowup","event_title":"Buckmaster and Alpöge, using Sol and Claude, prove smooth-forced blowup for Euler, IPM, and Boussinesq"},{"id":"5a05a004-c85c-43f2-b1be-a125f6503695","event_id":"49d392c2-d19e-4006-84de-0e821cb42c7e","model_id":null,"previous_score":"0.000000","new_score":"0.236700","previous_factors":{},"new_factors":{"evidence":0.7,"base_score":3,"durability":0.3,"attribution":0.85,"realization":0.45},"change_reason":"Imported events under methodology 0.2.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T00:37:09.904Z","event_slug":"aisi-mythos-5-unsanctioned-live-internet-2026","event_title":"Mythos 5 targets real people and a public GitHub project during a UK AISI cyber eval"},{"id":"2ad7047f-134c-4c99-a003-92cee07b0362","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"1.936000","new_score":"2.710400","previous_factors":{},"new_factors":{"flags":[],"evidence":0.7,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Independent METR/Redwood investigation, OpenAI alignment technical report, and May reconnaissance raise evidence.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T00:37:09.904Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"fdba9e18-f22c-44d9-a39b-39c2b4ad6acf","event_id":"8e3c228d-1df3-4e40-9d5e-6ad06a76bb17","model_id":null,"previous_score":"0.000000","new_score":"1.936000","previous_factors":{},"new_factors":{"evidence":0.5,"base_score":10,"durability":0.55,"attribution":0.85,"realization":0.8},"change_reason":"Imported events under methodology 0.1.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T23:22:50.724Z","event_slug":"openai-huggingface-eval-sandbox-escape","event_title":"OpenAI eval models exploit a Hugging Face zero-day and reach production systems"},{"id":"a62ee647-dd30-454f-aad4-45c8f25c6df3","event_id":"95a88b67-4c01-4574-b6d2-960cb3c2a1e3","model_id":null,"previous_score":"0.000000","new_score":"0.273000","previous_factors":{},"new_factors":{"evidence":0.35,"base_score":3,"durability":0.65,"attribution":0.8,"realization":0.5},"change_reason":"Imported events under methodology 0.1.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T22:56:05.431Z","event_slug":"gpt56-sol-chernoff-strong-log-concavity","event_title":"GPT-5.6 Sol proves Chernoff’s density is strongly log-concave"},{"id":"689971de-6c2a-4e78-a8d4-a4cb15d87d59","event_id":"2ea71b1c-1f22-41e4-be71-61dd65635475","model_id":null,"previous_score":"0.000000","new_score":"0.375400","previous_factors":{},"new_factors":{"evidence":0.5,"base_score":3,"durability":0.55,"attribution":0.65,"realization":0.7},"change_reason":"Imported events under methodology 0.1.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T22:56:05.431Z","event_slug":"gpt56-sol-short-superpermutations","event_title":"GPT-5.6 Sol / Codex produces shorter superpermutations for n=8, 9, and 10"},{"id":"2f76401c-fb48-4610-8c29-99cda8e4e835","event_id":"a5c7f692-39a0-4858-84cd-c8d4ebfb6dfd","model_id":null,"previous_score":"0.000000","new_score":"0.303200","previous_factors":{},"new_factors":{"evidence":0.35,"base_score":3,"durability":0.7,"attribution":0.75,"realization":0.55},"change_reason":"Imported events under methodology 0.1.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T22:56:05.431Z","event_slug":"gaussian-moments-conjecture-counterexamples","event_title":"Long, using Sol Pro and Fable 5, posts low-degree Gaussian Moments counterexamples"},{"id":"f7266e10-6ced-4530-980b-943cbf77f2a3","event_id":"5b65e7ad-7b20-42a1-9671-9b6942ee7c2a","model_id":null,"previous_score":"0.000000","new_score":"1.130700","previous_factors":{},"new_factors":{"evidence":0.5,"base_score":10,"durability":0.75,"attribution":0.55,"realization":0.6},"change_reason":"Imported September 2026 mathematics events under methodology 0.1 credit-unit rules.","trigger_type":"editorial_import","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-13T21:33:49.339Z","event_slug":"buckmaster-alpoge-smooth-forced-fluid-blowup","event_title":"Buckmaster and Alpöge, using Sol and Claude, prove smooth-forced blowup for Euler, IPM, and Boussinesq"}]}}