Skip to content
NeurIPS 2026 planner

Poster · Paris

More Value per Key: Asymmetric Sparse Attention for Faster LLM Decoding

When
Wed 9 Dec 2026, 17:00–19:00 CET
Where
Paris Poster Hall
Session
Paris Poster Session 2
Authors
Noam Elata, Itay Lamprecht, Mikey Shechter, Daniel Ohayon, Itay Hubara, Daniel Soudry
Organisations
Technion, Crusoe, Habana Labs
QRS classification
Attention and transformers, Language models

Open in scheduleOfficial page