Running Chasing the Counting Manifold in Open LLMs π We repeat some of their experiments on open models
Running Chasing the Counting Manifold in Open LLMs π We repeat some of their experiments on open models
Sanity Checks for Sparse Autoencoders: Do SAEs Beat Random Baselines? Paper β’ 2602.14111 β’ Published 4 days ago β’ 55
Running 86 Unlocking On-Policy Distillation for Any Model Family π 86 Visualize on-policy distillation for any model family
Running 3 Reproduction of the Anthropic's Counting Manifold π 3 We repeat some of their experiments on open models
T-pro 2.0: An Efficient Russian Hybrid-Reasoning Model and Playground Paper β’ 2512.10430 β’ Published Dec 11, 2025 β’ 115