Machine Learning

RLM Layer for Handwriting Correction

An application of Recursive Language Model (RLM) scaffolding to handwriting correction. Instead of correcting each page in isolation, the whole document's OCR output is loaded into a REPL-style environment and processed as one coherent body of text.

An application of Recursive Language Model (RLM) scaffolding to handwriting correction. Instead of correcting each page in isolation, the whole document's OCR output is loaded into a REPL-style environment and processed as one coherent body of text.
RLM Ollama Qwen RAG CPU
Full Overview View Source

Related Projects

Machine Learning

Cross-Machine OCR Escalation

Machine Learning

Graph-Based Agent Memory (Honcho)

Machine Learning

Handwriting Recognition Pipeline