Polylog
Sign inSubscribe
← The Polylog AI Intelligence Brief

Morning Edition · Wednesday, July 15, 2026Published at 1:45 AM EDT · New York

A New Linear-Attention Variant Adds Phase-Controlled Delta Memory to Fix Long-Context State Tracking

Semidirect Fourier Delta Attention keeps a fixed recurrent state instead of a growing key-value cache, using constructive chunk-WY kernels to preserve exact state tracking that linear attention normally loses.

Save
A New Linear-Attention Variant Adds Phase-Controlled Delta Memory to Fix Long-Context State Tracking

A new paper, Semidirect Fourier Delta Attention, targets the central weakness of linear attention. Linear attention replaces softmax attention's growing key-value cache with a fixed-size recurrent state, which bounds memory and cost, but th…

Continue the AI Intelligence Brief

Track frontier labs, chips, export controls, model releases, regulation, and AI infrastructure.

  • 5 AI intelligence signals a day
  • Frontier labs, compute, and chips
  • Model releases and AI infrastructure
  • Source-grounded analysis with confidence labels

The Global Intelligence Brief stays free.

Subscribe for $19/moAlready a member? Sign in

Part of a tracked trend

The Inference-Cost Efficiency Race

Techniques that cut tokens generated and KV-cache memory per query will keep compressing the marginal cost of serving reasoning models, making inference efficiency a recurring competitive axis alongside raw capability.

More from this edition