Mastodawn

Shawn Simister Jul 9, 2023

I've got prompts from 4 related papers marked up now. You can try them out here: explainprompt.com

Shawn Simister Jul 8, 2023

Working on a little side project to visualize prompting techniques like Chain of Thought. I like the idea of being able to break down a paper into bite-sized pieces and then step through the prompt like an interactive debugger.

Shawn Simister Dec 20, 2022

Did you know that you can use the logit_bias parameter on the #gpt3 API to feed in a specific set of tokens and see which one is the most probable next token

Show thread

Shawn Simister Dec 19, 2022

Also enjoyed:

A projection-mapped diffusion model with interactive hand-tracking.

A diffusion model drawing app on iPad with Apple pencil support.

A GPT-3 / DALL-E RPG game which suggests actions but also allows the player to interact with natural language.

An interactive tool for ML-powered microscopy image analysis.

#aihackweek

Shawn Simister Dec 19, 2022

#aihackweek demo day had so many great projects. Here are some of my favorites:

A #nocode tool which helps users build data transformation tools from natural language.

Adrenaline IDE: An AI-first code editor which can automatically debug runtime errors.

Motif.land: Interactive docs authoring with language model-powered refactoring.

The UI for AI: A graph-based editor for world building combining text and image models

Shawn Simister Dec 16, 2022

Using Blender to render a depth map and then feeding the depth map into #StabelDiffusion2 to render textures. Those textures then get projection mapped back onto the 3D model in Blender 🤯. https://github.com/carson-katri/dream-textures

GitHub - carson-katri/dream-textures: Stable Diffusion built-in to Blender

Stable Diffusion built-in to Blender. Contribute to carson-katri/dream-textures development by creating an account on GitHub.

GitHub

Shawn Simister Dec 16, 2022

Love this notebook by Peter Norvig analyzing the strengths and weaknesses of the code generated by DeepMind’s AlphaCode https://github.com/norvig/pytudes/blob/main/ipynb/AlphaCode.ipynb

pytudes/AlphaCode.ipynb at main · norvig/pytudes

Python programs, usually short, of considerable difficulty, to perfect particular skills. - pytudes/AlphaCode.ipynb at main · norvig/pytudes

GitHub

Shawn Simister Nov 28, 2022

Surprising lack of Googlers at this bar #NeurIPS2022 #NewOrleans

Shawn Simister Nov 24, 2022

Stable Diffusion 2.0 is out and it comes with the ability to synthesize images conditioned on prompts and depth maps making it easier to maintain consistency between images of the same subject https://github.com/Stability-AI/stablediffusion#depth-conditional-stable-diffusion

GitHub - Stability-AI/stablediffusion: High-Resolution Image Synthesis with Latent Diffusion Models

High-Resolution Image Synthesis with Latent Diffusion Models - Stability-AI/stablediffusion

GitHub

Shawn Simister Nov 23, 2022

Steps towards prompt-based creation of virtual worlds
https://arxiv.org/abs/2211.05875

Using the OpenAI Codex model to build an interactive Holodeck

Steps towards prompt-based creation of virtual worlds

Large language models trained for code generation can be applied to speaking virtual worlds into existence (creating virtual worlds). In this work we show that prompt-based methods can both accelerate in-VR level editing, as well as can become part of gameplay rather than just part of game development. As an example, we present Codex VR Pong which shows non-deterministic game mechanics using generative processes to not only create static content but also non-trivial interactions between 3D objects. This demonstration naturally leads to an integral discussion on how one would evaluate and benchmark experiences created by generative models - as there are no qualitative or quantitative metrics that apply in these scenarios. We conclude by discussing impending challenges of AI-assisted co-creation in VR.

arXiv.org