Surya Narreddi
surya.website · 870 words · saved by 16 readers
I'm a designer and creative technologist based in Brooklyn, NY.
Training AI to Paint with Code When you make an image with an AI model, the only way to participate is the prompt. You cannot edit the image directly. To change anything you go back to the model and prompt again. That limitation is what started this project. My friend Cameron and I trained a language model to make images by writing code, using reinforcement learning. The code is the artefact, and the code is editable. You can change what the model produced more granularly without going to prompt. The deeper question this project asks is how to do reinforcement learning on creative and…
saved by
- Anson Yu
- Tazik Sh
- Ratan Kaliani
- Rishi Kothari
- Benedict Neo
- Varun Goyal
- Ebaad Rehman
- Shubhaankar Sharma
- Emily Zhang
- Adani Abutto
- Akshaya Dinesh
- Aarya Gadekar
related reading
- Training a coding model to paint watercolours with TRL and OpenEnvhuggingface.co
- Improving Composer through real-time RL · Cursorcursor.com
- Illustrating Reinforcement Learning from Human Feedback (RLHF)huggingface.co
- State of RL for reasoning LLMs | A. Weersaweers.de
- RLHF: Reinforcement Learning from Human Feedbackhuyenchip.com
- Deep Reinforcement Learning: Pong from Pixelskarpathy.github.io
- AI Design Field Guideaidesignfieldguide.com
- My first Vibe RL experiencesriraam.me
- Harness design for long-running application developmentanthropic.com
- Learning Beyond Gradientstrinkle23897.github.io
- Painting With Concepts Using Diffusion Model Latentsgoodfire.ai
- Features as Rewards: Using Interpretability to Reduce Hallucinationsgoodfire.ai