In recent years, transformer-based models have revolutionized deep learning, particularly in sequence modeling.

The code is available here.

Exploiting symmetry inherent in data can significantly improve the sample efficiency of alearning procedure and the generalization of learned models.

Source code is made available at https://github.com/ymLeiFDU/LICO.

In this work, we investigate how an adversary can exploit instruction tuning by injecting specific instruction-following examples into the training data that intentionally changes the model's behavior.