Plant DoctorPlant Doctor
Plant recognition, care suggestions, and plant problem guidance.
I am HotPig, an independent AI app developer. This site collects my apps, web tools, vlogs, and technical writing.

The first LLM course note: tokenization, embeddings, Transformer layers, probabilities, and autoregressive generation.
Plant recognition, care suggestions, and plant problem guidance.
Personal apps that are live or launching, with regional download entries.
Plant recognition, care suggestions, and plant problem guidance.
Turns recordings, audio, and video into structured meeting notes and work documents.
A reader for PDFs, documents, materials, excerpts, annotation, summaries, and knowledge organization.
A structured personality assessment for self-understanding and personal observation.
Voice recording, AI transcription, content organization, and card generation.
Video notes about product building, AI learning, and indie development will live here.
Writing is the core of this personal site. Articles are organized into LLM, development, and SOP columns.
The first LLM course note: tokenization, embeddings, Transformer layers, probabilities, and autoregressive generation.
The first LLM course note: tokenization, embeddings, Transformer layers, probabilities, and autoregressive generation.
Why an LLM answers a prompt: conditional probability, pattern activation, and hallucination.
How LLM parameters improve through forward pass, loss, backpropagation, gradient descent, and alignment stages.
Foundation models differ by architecture, data, context, post-training, openness, cost, and product fit.
Convergence is optimization, not truth. Generalization determines whether the model works on unseen problems.
A practical learning roadmap for independent developers: understand the structure first, then build projects with APIs, RAG, agents, evals, and production workflows.
Personal positioning, contact, language versions, and filing information.
I keep turning learning records, product thinking, and project progress into written notes and working products.
For support, privacy questions, App Store review, or formal collaboration.
support@hotpig.com.cnFor WeChat, please include your reason.
Please include your reason when adding WeChat.

If an article helped you, you can support with a donation QR code.
