2026-05-02
Can a simple agentic LLM achieve similar regression model selection results as exhaustive GridSearchCV — while using far fewer model evaluations? I built a Claude-powered agent that profiles datasets and picks models, then benchmarked it head-to-head against brute-force search and Bayesian optimisation across five datasets.
Can a simple agentic LLM achieve similar regression model selection results as exhaustive GridSearchCV — while using far fewer model evaluations? I built a Claude-powered agent that profiles datasets and picks models, then benchmarked it head-to-head against brute-force search and Bayesian optimisation across five datasets.
Read more →
2026-04-29
How demand-weighted and affinity-aware item placement reduces pick-agent travel distance by up to 30% compared to random slot assignment — tested across both small and large warehouse layouts with 1,000 orders each.
How demand-weighted and affinity-aware item placement reduces pick-agent travel distance by up to 30% compared to random slot assignment — tested across both small and large warehouse layouts with 1,000 orders each.
Read more →
2026-04-28
Training a Double Deep Q-Network agent to play Super Mario Land — covering environment wrappers, replay memory, target networks, and reward shaping.
Training a Double Deep Q-Network agent to play Super Mario Land — covering environment wrappers, replay memory, target networks, and reward shaping.
Read more →
2026-04-27
A placeholder post to test the blog layout and markdown rendering.
A placeholder post to test the blog layout and markdown rendering.
Read more →