Questions concerting encoding of grid: what "tiled" stands for?

[thua919]

Hi, Thanks so much for this great work!

I have a small question, the grid encoding part allows three kinds of storage: hashtables, dense, or tiled. I am not completely sure about what "tiled storage" might mean and where to use that: I am wondering if this could be a kind of plane, and we can harness it for, e.g. Tri-plane implementation NeRF? In that case how to specify 3 planes accordingly?

embed = tcnn.Encoding( n_input_dims=input_dim, encoding_config={ "otype": "Grid", "type": "tiled", "n_levels": n_levels, "n_features_per_level": level_dim, "base_resolution": base_resolution, "per_level_scale": per_level_scale, "interpolation": "Linear"}, dtype=torch.float )

I am looking forward to your kind response, many thanks!