Extending Anthropic's Agent Workflows with Recursive Planning
Custom agent design using Google's new ADK framework.
Experiments in Constrained LLM Decoding
Writing with only the 1,000 most common words in the English language.
AIMO Progress Prize 1
Solving national-level math challenges using AI models.
Generative AI Interfaces
Seven recurring design elements found in Generative AI applications and experiments.
Bletchley Park's Legacy in Code
Reimplementing the Enigma and Bombe machines using Python.
Adapting Gymnasium to Find DeepRacer Racing Lines
Train a PPO agent from Stable Baselines3 to master the Gymnasium CarRacing environment.
Driving Trajectory Extraction from Image Sequences
Learn how to train the DeepVO visual odometry model on the KITTI dataset.
Handwritten Equation Recognition with Transformers
Learn how to turn handwritten math into LaTeX using Transformers and the CROHME 2023 dataset.
Deriving Blackjack’s Basic Strategy using Reinforcement Learning
Tabular solution methods from reinforcement learning are used to derive Blackjack’s basic strategy.
Revisiting the Algorithm that Changed Horse Race Betting
Explore Bill Benter’s iconic horse betting strategy, enhanced with modern code and contrasted against three decades of evolving data.
Advanced Horse Race Tactics using Trakus Coordinate Data
The impact of drafting percentage, path efficiency, race strategy and speed fluctuation on horse race tactics.