Diffusion models have achieved remarkable progress in the field of image generation due to their outstanding capabilities.

In this work, we propose an innovative test-time poisoned sample detection framework that hinges on the in-terpretability of model predictions, grounded in the semantic meaning of inputs. We contend that triggers (e.g., infrequent words) are not supposed to fundamentally alter the underlying semantic meanings of poisoned samples as they want to

We present a comprehensive study of task arithmetic in vision-language models and show that weight disentanglement is the crucial factor that makes it effective. This property arises during pre-training and manifests when distinct directions in weight space govern separate, localized regions in function space associated with the tasks.

However, the nonlinearity of risk measures makes it challenging to achieve convergence and optimality.

Public benchmarks play an essential role in the evaluation of large language models.