Ostris pre-trains "Primal VAE": 2x spatial downsampling to a single channel
Crazy-Repeat-2006 · reddit · 2026-10-26
Ostris is pre-training a new VAE concept called Primal VAE: only 2x spatial downsampling down to 1 channel. Patching the latent by 4 yields 8x at 16 channels — equivalent to most modern VAEs — without deep compression. Shared reconstruction samples look close to the input; the approach could reshape how image-generation models compress latents.
Related event: Ostris Pretrains Primal VAE, a Radical Single-Channel Approach(3 posts)→
More from Multimodal
- 10 practical ways to run Astra computer use on Premiere Pro for video editing — beechinour · 2026-10-27
- MiniMax H3 runs fully on a Galaxy S26: 5-second video with audio in ~7 minutes — BidFunny4862 · 2026-10-27
- Krea ships Premiere Pro plugin to generate video and SFX inside the timeline — lmoroney · 2026-10-27
- Wan2GP lands on Pinokio: 1-click AI video on 6GB GPUs — cocktailpeanut · 2026-10-27
- WanGP v18.81 adds Tencent's Hy-Nomas world model, playable on 8GB VRAM — cocktailpeanut · 2026-10-27
- Butter.video wins over skeptical creator with hundreds of tasteful timeline 'blocks' for video effects — beechinour · 2026-10-27