Self Information Update for Large Language Models through Mitigating Exposure Bias

Open in new window