6 papers
On the Convergence of Moral Self-Correction in Large Language Models
Guangliang Liu, Haitao Mao, Bochuan Cao +4
Large Language Models (LLMs) are able to improve their responses when instructed to do so, a capability known as self-correction. When instructions provide only a general and abstr…
No Free Lunch for Defending Against Prefilling Attack by In-Context Learning
Zhiyu Xue, Guangliang Liu, Bocheng Chen +2
The security of Large Language Models (LLMs) has become an important research topic since the emergence of ChatGPT. Though there have been various effective methods to defend again…
Smaller Large Language Models Can Do Moral Self-Correction
Guangliang Liu, Zhiyu Xue, Xitong Zhang +2
Self-correction is one of the most amazing emerging capabilities of Large Language Models (LLMs), enabling LLMs to self-modify an inappropriate output given a natural language feed…
Towards Understanding Task-agnostic Debiasing Through the Lenses of Intrinsic Bias and Forgetfulness
Guangliang Liu, Milad Afshari, Xitong Zhang +5
While task-agnostic debiasing provides notable generalizability and reduced reliance on downstream data, its impact on language modeling ability and the risk of relearning social b…
On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept
Guangliang Liu, Haitao Mao, Bochuan Cao +5
Large Language Models (LLMs) are able to improve their responses when instructed to do so, a capability known as self-correction. When instructions provide only the task's goal wit…
PAC-tuning:Fine-tuning Pretrained Language Models with PAC-driven Perturbed Gradient Descent
Guangliang Liu, Zhiyu Xue, Xitong Zhang +2
Fine-tuning pretrained language models (PLMs) for downstream tasks is a large-scale optimization problem, in which the choice of the training algorithm critically determines how we…