Early run on a Factorio agent benchmark and env I’m building (30 mins at 120x), and a textbook demonstration of emergent reward hacking.
(Using DeepSeek V4.1 Flash) It looks active, but it collapsed into a local optimum almost immediately: shuffling coal, hand-smelting plates,
Today's update for
1. Custom alerts are now live. You can configure alerts for balance thresholds, model changes, and more. 2. Deepseek's cheapest and smartest model is now live: deepseek-v4.1-flash. 3. Image generation is now live. Please let us know
DeepSeek 4.1 Flash vs. GPT-6 Astra + Higgsfield in a domino simulation
DeepSeek Flash 4.1 speed is addicting
About to take a nap, and I just set off a prompt to build out a tower defense game. Termius/Herdr/OMP
You have reached the end of the archive
All of deepseek