Tag
The OpenAI Developers account posted a one-minute video of Cognition co-founder Walden Yan on September 15. Inside Devin, GPT-6 Astra backs up the claim that code works with tests, cutting the amount of code people have to read themselves.
The Tencent Hunyuan team released the EvolveScaler paper and project page on September 15, an information-evolution data framework. On data built by defining state in code first and then rendering it as natural language, the median avg@5 across 14 models fell to 11.3 on the longest tier.
Specific released Real-SWE on September 12, a benchmark that measures AI coding models on private production codebases licensed from real companies. Fable 5.1 led with a 38.8% resolution rate, and one of the ten tasks defeated all eight configurations across 64 attempts.
Fugu Max routes each task to the leanest model that can solve it and undercuts rival output pricing by 40 to 60 percent. Fugu Ultra v2 posted its best scores with Fable 5.1 and GPT-6 Astra kept out of its pool.
Four days after receiving a "critical" cyber risk rating, OpenAI showed the model solving hacker puzzles using parallel agents
OpenAI calls Astra the start of the AGI era, but safety experts worry its reasoning can't be fully read
In a new test that hands AI a year of running an online store, GPT-5.6 Sol took first place while an older Alibaba model went bankrupt four times
A model trained on a European cohort actually lost accuracy as the target population's sample size grew
OpenAI begins rolling out Astra, a model that handles computer tasks on your behalf, starting today
Across 7.1 million verification images, a training-free discriminator matched or beat existing detectors
Built with HHMI Janelia, the map covers 166,000 neurons and 125 million synapses, making it the largest connectome ever produced
Of 1,320 mini-protein binders Claude designed, 354 actually bound — hitting 14 of 15 targets. One target that stumped human competitors, who managed only a 3.7% hit rate, yielded 40% for Claude. But picking winners before synthesis is still the hard part.
