megadragon9
- Karma
- 73
- Created
- 12 years ago
About
Getting back on the right track.Recent Submissions
- 1. ▲ Show HN: Auto-train the harness, not the LLM. cross-model, cross-benchmark gains (github.com)
- 2. ▲ Freeze the model, train the harness: gains transfer across LLMs and benchmarks (henrypan.com)
- 3. ▲ Training Agent Harness Like Training a ML Model (henrypan.com)
- 4. ▲ Show HN: Freeze the Model, Train the Harness (github.com)
- 5. ▲ Train a Harness to improve model/env-agnostic capabilities with PyTorch-like API (github.com)
- 6. ▲ GPT-2 124M checkpoint pre-trained on OpenWebText 27.5B tokens (github.com)
- 7. ▲ Self-Improving Harness Is an Experiment Design Problem (henrypan.com)
- 8. ▲ Show HN: What 1k Harness Experiments Taught Me About Self-Improving Agents (henrypan.com)
- 9. ▲ What 1k Harness Experiments Taught Me About Self-Improving Agents (henrypan.com)
- 10. ▲ How a Deep Learning Library Enables Learning (henrypan.com)