I don’t think Lreconstruction is very informative here, as it’s highly impacted by the input batch. Both the raw Lreconstruction and Lclean have large variances at different verification steps, and since we mainly care about how good our reconstruction is compared with the original, I think the reconstruction score is good as is. I also don’t follow why the noisiness of L0 leads to showing Lreconstruction.
I don’t think Lreconstruction is very informative here, as it’s highly impacted by the input batch. Both the raw Lreconstruction and Lclean have large variances at different verification steps, and since we mainly care about how good our reconstruction is compared with the original, I think the reconstruction score is good as is. I also don’t follow why the noisiness of L0 leads to showing Lreconstruction.