Adobe Research and UCLA propose the Tev model
The story so far
Adobe Research and UCLA have proposed the Tev model, a 27B parameter-scale open-source model that matches the performance of GPT-6 Astra and Claude Opus 5 in long-horizon reasoning tasks. Long-horizon agents are considered key to automated scientific discovery and recursive self-improvement (RSI), but open-source models significantly lag in long-horizon tasks, lacking training data and effective training recipes. The Tev model enhances agent capabilities by constructing an 'answer world' and designing long-horizon questions.
AI-written from the reports · Updated 27 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- SyncedAdobe Research and UCLA jointly propose the Tev model, an open-source 27B model that matches the performance of GPT-6 Astra and Claude Opus 5 on long-range reasoning tasks.
The Tev model proposed by Adobe Research and UCLA matches the performance of GPT-6 Astra and Claude Opus 5 on long-range reasoning tasks. This model is an open-source 27B parameter model that enhances agent capabilities by constructing a 'world of answers' and designing long-range questions.