Weird output from Kilosort 2

[shruticn]

I have been running Kilosort on data that I collected using an Intan system. I have generally been able to spike sort with this data, but for some of my recent files I get this weird looking output from Kilosort when I load it in phy. The raw data does not look particularly noisy (except for some periodic movement-related artifact) so I'm wondering what might be going on with the sorting?
! rawdata

Thank you!