Many thanks for sharing the paper list. Could you please add our new paper as follows?
Title: Scaling Video-Language Models to 10K Frames via Hierarchical Differential Distillation
Venue: ICML 2025
Paper: https://arxiv.org/abs/2504.02438
Code: https://github.com/steven-ccq/ViLAMP