Show HN: Auto-train the harness, not the LLM. cross-model, cross-benchmark gainsgithub.com·4 pts·megadragon9·0
Freeze the model, train the harness: gains transfer across LLMs and benchmarkshenrypan.com·2 pts·megadragon9·0
Train a Harness to improve model/env-agnostic capabilities with PyTorch-like APIgithub.com·3 pts·megadragon9·0
Show HN: What 1k Harness Experiments Taught Me About Self-Improving Agentshenrypan.com·3 pts·megadragon9·0
Understanding ML: Creating a PyTorch-Inspired Deep Learning Library from Scratchhenrypan.com·1 pts·megadragon9·0
Show HN: ML Library created by Python+NumPy (trains CNNs to a toy GPT-2)github.com·1 pts·megadragon9·0
I Built a Deep Learning Library from Scratch Using Only Python, NumPy and Mathhenrypan.com·3 pts·megadragon9·0
Show HN: From-Scratch ML Library (trains models from CNNs to a toy GPT-2)github.com·1 pts·megadragon9·0