IO-aware attention for long sequences on a single GPU

IO-aware attention tiling cuts HBM traffic on long sequences and composes with a streaming softmax reduction.

structural read; full verdict pending

Holds when: single GPU; long sequences

Evidence ledger

• S1

  Present: An IO-aware attention tiling method reduces HBM traffic for long-context decode. Complements streaming softmax reductions.

  Present: https://github.com/example/io-attention

• S2

  Present: arXiv subjects: machine learning.

• S3

  Present: Official repository listed.

  Not available yet: Container dry-build is not available yet

• S4

  Absent: No distinct-team reimplementation with matching numbers

• S5

  Absent: No Hugging Face model card cites this paper.

• S6

  Present: OpenAlex lookup recorded. Context classifier excluded from publish until 500 labels.

• S7

  Absent: No OpenReview decision attached

• S8

  Absent: No named-account repro thread

• S9

  Absent: No independent harness match

• S10

  Not available yet: S10 is not available yet

• S11

  Not available yet: S11 is not available yet

• S12

  Present: S12 band factor only. Prestige is not substantive evidence.

• S13

  Absent: No public dataset link

• S14

  Absent: No merged serving-framework PR

• S15

  Absent: No Alora audit attached at extract

• S16

  Not available yet: S16 is not available yet