same-hardware latency framing (138x/65x re-timed) + replace stale p>0.75 with effect-size framing

fix VAE latent dims: 32 tokens x 32-d (1,024-d total), not 32x1024

update card: SMC paper accepted + renamed CloudDiffusion; correct GT params (w=17, 0.1m), alpha_bar/SNR, vae_v3 size + load instructions

rebrand to ScaffDiff: update headline numbers to 6-seed (0.968 ± 0.194), 69-72% vs baselines, paper title

add mixed-scaffold fine-tuned teacher (3 epochs from teacher_v2gt, 30% lidar-cropped scaffolds, x0-chamfer loss; cd² 0.727 m² paired scaffold-free)

rewrite README around RA-L scaffold-dominant + FT-win and SMC VAE+v2GT papers; deprecate old modality-agnostic / distillation framing

fix results: correct student CD, add modality agnosticism finding

Upload README.md with huggingface_hub

Upload README.md with huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload README.md with huggingface_hub

Upload folder using huggingface_hub

initial commit