Runtime steering vectors for local models that work with MLX and PyTorch through PTS, with the data behind them.
AI & ML interests
Local LLMs
Recent Activity
Runtime steering vectors for local models that work with MLX and PyTorch through PTS, with the data behind them.
Small subsets of agent benchmarks for comparing systems at a fraction of the cost. Each card says how closely it tracks the full benchmark.
Synthetic datasets built with DataInventor: describe the data you want and get training rows back.
Open models that answer yes/no, choice and score questions with probabilities, plus typed-decisions, a public benchmark to compare them.