Jianhong Tu

zero-shot learning (1) benchmark evaluation (1) transfer learning (1) knowledge transfer (1) factual accuracy (1) multimodal learning (1) constrained optimization (1) instruction tuning (1) information acquisition (1) vision-language model (1) zero-shot generalization (1) long-horizon planning (1) multimodal language model (1) long-form generation (1) large language model (1) travel planning (1) text-only training (1) verifiable evaluation (1) real-world complexity (1) verifiable assessment (1)