Pith. sign in

Proceedings of ICLR , year=

Efficient streaming language models with attention sinks , author=

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
browse 1 citing papers

fields

cs.CL 1

years

2026 1

verdicts

CONDITIONAL 1

representative citing papers

Hybrid Gated Attention

cs.CL · 2026-08-12 · conditional · novelty 5.0

HyGA, a hybrid gated attention with input, output, and cross-head gates plus low-rank compression, improves LLM training loss and downstream tasks over Gated Attention.

citing papers explorer

Showing 1 of 1 citing paper.

  • Hybrid Gated Attention cs.CL · 2026-08-12 · conditional · none · ref 5

    HyGA, a hybrid gated attention with input, output, and cross-head gates plus low-rank compression, improves LLM training loss and downstream tasks over Gated Attention.