Pith. sign in

Compute-optimal Context Size , b

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

1 Pith paper citing it

fields

cs.LG 1

years

2025 1

verdicts

CONDITIONAL 1

representative citing papers

Scaling Context Requires Rethinking Attention

cs.LG · 2025-07-06 · conditional · novelty 4.0

On 64k-token natural language training, power attention with degree 2 achieves lower loss per FLOP than both softmax attention and existing linear attention, and its GPU kernels run faster than Flash Attention at long context.

citing papers explorer

Showing 1 of 1 citing paper.

  • Scaling Context Requires Rethinking Attention cs.LG · 2025-07-06 · conditional · none · ref 4

    On 64k-token natural language training, power attention with degree 2 achieves lower loss per FLOP than both softmax attention and existing linear attention, and its GPU kernels run faster than Flash Attention at long context.