PhD @ Max Planck Institute. Prev: Post-Training @MistralAI
Past @iiit_hyderabad@VectorInst. Interested in better evals, forecasting, and open-endedness.
FutureSim cleanly differentiates the performance of the agents on their skill of forecasting with GPT-5.5 dominating.
We find open-weights models are significantly overconfident in their predictions, performing worse than the baseline of not submitting any prediction. (4/5)