Launching today
Lensa ML
108 interactive ML explainers, from calculus to LLM serving
0 followers
108 interactive ML explainers, from calculus to LLM serving
0 followers
Free interactive explainers for machine learning and the math under it. Each topic is six steps, and the sliders drive the real formulas. It runs from calculus to transformers, plus LLM serving and three system design case studies. No account needed.
Free
Launch Team
ArticosYour audience, ready to react to your messaging in minutes. Try free
Promoted
Hi Product Hunt,
Lensa ML is a free site of interactive explainers for machine learning and the math under it. Each topic is six short steps, and the sliders run the real math: raise the learning rate and gradient descent starts to overshoot, add tokens and the KV cache grows.
There are 108 explainers. They start with derivatives, linear algebra and probability, go through classic ML and deep learning, and end with transformers and how LLMs are trained and served (sampling, quantization, FlashAttention, speculative decoding, distributed training). Three system design case studies let you set targets for latency, throughput and cost, then watch the GPU count, recall and cost per query change.
It's free with no account. Progress stays in your browser, and each explainer has a written version under the visuals with links to the papers.
Good places to start:
- Gradient descent: https://lensaml.com/explainers/g...
- KV cache: https://lensaml.com/explainers/k...
- Build a RAG system: https://lensaml.com/system-desig...
Corrections are welcome, especially where an explanation is wrong or a simplification goes too far.