(In case it helps: I'd also expect a public per-sample entry point like gradient one(...) on the…

(In case it helps: I'd also expect a public per-sample entry point like gradient one(...) on the…: a task in MiMo-V2.6-RL-oss: Agentic RL Environments (MiMo RL release). Hi, I have tested the newest version of Foolbox and it seems like it can't handle parallel batch attacks with estimated…