Semantha — A/B prompt comparison over SAE features
Compare two prompts through Neuronpedia SAE features
Upscale images 4× with AuraSR
Generate detailed captions for any image
Building and scaling RL environments for LLM training
pixel-perfect monocular depth estimation
SOTA real-time object detection model
Real-time object detection w/ 🤗 Transformers.js
Amplify motion in videos by adjusting frames and brightness
Next-Gen High-Resolution 3D Model Generation
Sparse-view SFM-free Gaussian Splatting in Seconds
Generate 3D models from images
mcp_server & FLUX 4-bit Quantization + Enhanced
Create 3D models from images using depth estimation
Execute custom code provided in environment variables
Generate detailed prompts from an image
Upscale low‑resolution images to high‑resolution versions