Pinned Loading

1. world-model-harness world-model-harness Public

   World-model-as-a-harness for simulating AI agent environments

   Python 63 3

2. openclaas openclaas Public

   Continual Learning as a Service

   Python 57 5

3. autoresearch-distillation autoresearch-distillation Public

   Open-source repo for applying continual learning to autoresearch with SDPO and other RL algorithms. Current example: Qwen3-14B achieves 1.023 val_bpb (−3.1%), surpassing the original Karpathy agent…

   Python 9