← back to paper

Review history

arxiv: 2608.03083 · 2 revisions

GSTEP: Global Spatio-Temporal Density-Driven Visual Token Pruning for Efficient Video Large Language Models

1. 2026-08-15 CONDITIONAL MODERATE v1.4.0-alltime-deepseek-medium novelty 6.0
   120266 ms 16163 in 14187 out 2026-08-15T14:52:54.006764+00:00
2. 2026-08-08 CONDITIONAL MODERATE v1.4.0-alltime-deepseek-medium novelty 5.0
   106352 ms 16161 in 12007 out 2026-08-08T00:56:37.666747+00:00