Peter’s Notes

From A/B to RL (3/3): MENACE and Delayed Rewards Learning state-dependent policies with MENACE, matchboxes, and beads.

From A/B to RL (2/3): Bandits and Thompson Sampling Learn and adapt online instead of choosing once at the end.

From A/B to RL (1/3): Bayesian A/B Testing Before policies learn, experiments assign.

Beta Priors and Pólya Urns for Binary Decisions Why the Beta prior naturally matches sequential binary feedback.

Beta distribution: Probabilities and binary data Many statistical problems start with an unknown probability.

Flow Matching: A visual introduction Flow Matching (FM) has become a prevalent technique to train a certain class of generative models.

ReAct agent with OpenAI function calling Simple example of using OpenAI Function Calling in a ReAct Loop .

The goal of this project was to investigate the capabilities of contemporary large language models (LLMs) such as GPT-3.5 and GPT-4 for discovering and chaining previously unseen APIs to execute tasks requiring multiple step-by-step actions.

The posts on this site are tagged according to their content.

About this blog I started this blog in 2015 from some notes I took in Jupyter Notebooks.