safetycritiquebearishExisting defenses for distributed learning typically address privacy and manipulation threats in isolation and are limited to specific paradigms or architecturesMachine Learning27 Jul 2026http://arxiv.org/abs/2607.02187v1