Temporal Gains, Spatial Costs: Revisiting Video Fine-Tuning in Multimodal Large Language Models

Temporal Gains, Spatial Costs: Revisiting Video Fine-Tuning in Multimodal Large Language Models explores This research explores the impact of video-based fine-tuning on multimodal large language models, revealing trade-offs in visual performance.. Commercial viability score: 4/10 in Multimodal Learning.