Devlog #5

Today i finally evaluated gpt 5.6-sol on the benchmark and found many things that we can change and tweak to improve the benchmark more
and its detailed trajectory for research use!
after that i cleaned up the codebase added docs images etc. for production release

Here’s a sneak peak of Gpt-5.6-sol walked path.