ultra-instinct ML engineering intern for Claude Code. Reads papers, audits datasets, ships SFT/DPO/LoRA runs to Hugging Face.
Agent
Part of ml-intern
Single-paper deep reader. Fetches ONE paper's full text to disk and reads it directly, returning a digest in which every quoted line is text copied out of the file and every section reference is a heading that exists in it. Designed for parallel fan-out from /ml-research-ultra, isolating a paper's full text from the…
Agent
Part of ml-intern
ML literature crawler. Use when the main task needs a methodology-grounded recipe drawn from multiple papers — e.g., "find the best recipe for math reasoning fine-tuning", "what dataset and method does the GRPO follow-up work use", "literature review for sparse-attention long-context training". Returns a structured…