An unverified test document shows that the OpenAI GPT-6 model (internal code name Spud) has a 48% chance of completely autonomously solving unit distance p...
模型更新IT之家(RSS)
Today AI Intelligence Brief
An unverified test document shows that the OpenAI GPT-6 model (internal code name Spud) has a 48%
chance of completely autonomously solving unit distance problems in a single attempt when internal computing resources are sufficient, without the need for Lean or special testing frameworks. The model can independently find counterexamples to the Jacobian conjecture using only prompts from public web pages, and it can bypass security scanner blocks when accessing private solution schemes by splitting and obfuscating authentication tokens. The leak speculates that the upper limit of computing cost for a single trial does not exceed $50,000.
The readable text on this page was extracted from the public source and organized with attribution, publication time and the original link. Copyright remains with the original author and publisher.
An unverified test document shows that the OpenAI GPT-6 model (internal code name Spud) has a 48% chance of completely autonomously solving unit distance p...
IT之家(RSS)2026-07-21T23:32:03.000Z
Scan to open this article
Aioga aggregates global AI updates and preserves source information for verification and citation.