← Search

Meijun Gao

2 accepted papers

2025

DAMO: Decoding by Accumulating Activations Momentum for Mitigating Hallucinations in Vision-Language Models

ICLR 2025poster

Large Vision-Language Models (VLMs) exhibit significant potential in multimodal tasks but often struggle with hallucinations—responses that are plausible yet visually ungrounded. In this work, we investigate the layer-wise prediction tendencies of VLMs and conduct an in-depth analysis of their decod…

Cited by 0SourcePDFScholar
2025

Layer-Level Self-Exposure and Patch: Affirmative Token Mitigation for Jailbreak Attack Defense

NAACL 2025long

As large language models (LLMs) are increasingly deployed in diverse applications, including chatbot assistants and code generation, aligning their behavior with safety and ethical standards has become paramount. However, jailbreak attacks, which exploit vulnerabilities to elicit unintended or harmf…