vix.ing
·
top
·
new
·
best
·
stats
·
spec
Haozhe Hu
WIDE: Boosting Adaptive LLM Inference via Token-level Dynamic Width Pruning
2026/07/30 by
Haozhe Hu
,
Hao Wu
,
Peiran Yin
+3
Computer Science
·
#cs.AI
#cs.CL
#cs.LG