Zixian Ma

vision-language model (4) benchmark evaluation (3) adversarial robustness (3) multimodal language model (3) natural language processing (2) visual question answering (2) text classification (2) compositional reasoning (2) adversarial evaluation (2) multimodal learning (2) adversarial training (2) textual adversarial attack (2) scene graph generation (1) visual reasoning (1) referring expression (1) video understanding (1) model robustness (1) expectation alignment (1) question answering (1) intrinsic reward (1)