### Bug Description @mmacklin isolated the issue down to `test_vbd_bending_cpu`:  > In the screenshot the weight_buckets vector has length=64, but bucket_idx=18446744073709551615. And the offending code: https://github.com/NVIDIA/warp/blob/2322b3b973c7cb235ceebb49b6b300df87a45cee/warp/native/coloring.cpp#L375 ### System Information _No response_