Paper Citation Record · LEDGER

Fast attention mechanisms: a tale of parallelism

As of 20 August 2026, this Paper Citation Record lists 17 of 17 outbound references and 1 inbound Pith citation observation for arXiv:2509.09001.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2509.09001 v1

Coverage vector

measured 17 of 17 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-04T20:03:11.488237Z

measured 18 of 18 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-20T06:33:59.587034+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-03T21:37:26.054535Z

measured 1 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: arxiv_reference, observed 2026-08-05T02:28:24.338817Z

Reference resolution

17 of 17 outbound references displayed

verified exact0
verified fuzzy0
unresolved17
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

0
arxiv_reference, observed 2026-08-05T02:28:24.338817Z

Outbound references

Observation 0f1e8470-3ddd-4b86-9020-07c67d8d11d7 · outbound

This paper cites The input feeded into EMA-transformer is distributed in theN tokens, and we need to transfer than into the first⌈N s ⌉tokens/machines to matchMachineIn 1.

Fast attention mechanisms: a tale of parallelism The input feeded into EMA-transformer is distributed in theN tokens, and we need to transfer than into the first⌈N s ⌉tokens/machines to matchMachineIn 1

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.457065Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0e356e5d-edec-4f04-8539-2d6a7f0b5471 · outbound

This paper cites After the local computation in each roundr, we need to communicate the messages fromMachineOutr toMachineIn r+1.

Fast attention mechanisms: a tale of parallelism After the local computation in each roundr, we need to communicate the messages fromMachineOutr toMachineIn r+1

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.459652Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9fd88597-255f-41eb-b97b-7233589b70cb · outbound

This paper cites Consider the family G with Prg∈G[g(x) = g(y)] ≤ 0.1 N 3, if ∥x−y∥> cr.

Fast attention mechanisms: a tale of parallelism Consider the family G with Prg∈G[g(x) = g(y)] ≤ 0.1 N 3, if ∥x−y∥> cr

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.453864Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5331295e-0e5e-4759-a794-9da91b19272c · outbound

This paper cites 32 For step 1, each machine only needs to send message to its neighbor machine: machinei sends message to machinei− 1, and this only takes 1 round.

Fast attention mechanisms: a tale of parallelism 32 For step 1, each machine only needs to send message to its neighbor machine: machinei sends message to machinei− 1, and this only takes 1 round

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.483375Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7fb2c500-9b9e-42d3-9899-557ec1594fc6 · outbound

This paper cites validity bit.

Fast attention mechanisms: a tale of parallelism validity bit

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.462237Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e5bc2661-6200-4c14-bead-106bda3b9d19 · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.465447Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bfc8912e-40c7-4732-8f38-11f122ad3a89 · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.468061Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d50e70c8-059a-4455-8796-6594b51ea772 · outbound

This paper cites The output embedding forqi is therefore X j:h′(kj )=h′(qi) exp(⟨qi, kj⟩)X j′:h′(kj′ )=h′(qi) exp(⟨qi, kj′⟩) ·v j.

Fast attention mechanisms: a tale of parallelism The output embedding forqi is therefore X j:h′(kj )=h′(qi) exp(⟨qi, kj⟩)X j′:h′(kj′ )=h′(qi) exp(⟨qi, kj′⟩) ·v j

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.470580Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 35ff3201-f943-4e63-9b8a-ced7364f731e · outbound

This paper cites In theℓ-th layer attention computation for each queryqi, Reformer computes X j∈fℓ(i) exp(⟨qi, kj⟩)X j′∈fℓ(i) exp(⟨qi, kj′⟩) ·v j.

Fast attention mechanisms: a tale of parallelism In theℓ-th layer attention computation for each queryqi, Reformer computes X j∈fℓ(i) exp(⟨qi, kj⟩)X j′∈fℓ(i) exp(⟨qi, kj′⟩) ·v j

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.473101Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 674bc6e7-b76c-48c4-847d-9feab1c8c92c · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.475927Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9b462122-3abd-40e5-9c9a-f1f535e2ede4 · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.478477Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cfad6d26-2ffa-45f2-a9be-3cdadfbcc53b · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.480929Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9927f75c-26cd-4cbf-838e-98865b468985 · outbound

This paper cites an unresolved cited work.

Fast attention mechanisms: a tale of parallelism Unresolved cited work

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.485823Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0b4b6cfe-9a7f-417b-a96b-eba2293780c8 · outbound

This paper cites Therefore, the total round isO( 1 ε ) + 2(ℓ + 1).

Fast attention mechanisms: a tale of parallelism Therefore, the total round isO( 1 ε ) + 2(ℓ + 1)

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.488237Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 81c44fe3-f049-4e39-9a8a-0134724c932a · outbound

This paper cites One-layer transformers fail to solve the induction heads task.

Fast attention mechanisms: a tale of parallelism One-layer transformers fail to solve the induction heads task

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.447661Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 96571bd3-f514-4cae-8968-c369a41e62e1 · outbound

This paper cites Nathan Linial.

Fast attention mechanisms: a tale of parallelism Nathan Linial

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.444053Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8fed9af1-48f0-4674-aed1-f86b4c2dd317 · outbound

This paper cites Kai Yang, Jan Ackermann, Zhenyu He, Guhao Feng, Bohang Zhang, Yunzhen Feng, Qiwei Ye, Di He, and Liwei Wang.

Fast attention mechanisms: a tale of parallelism Kai Yang, Jan Ackermann, Zhenyu He, Guhao Feng, Bohang Zhang, Yunzhen Feng, Qiwei Ye, Di He, and Liwei Wang

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-04T20:03:11.451157Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 5be28c95-e3e0-49fe-9b5a-71d17c4da13e · inbound

Scaling Laws for Grid-Based Approximate Nearest Neighbor Search in High Dimensions cites this paper.

Scaling Laws for Grid-Based Approximate Nearest Neighbor Search in High Dimensions Fast attention mechanisms: a tale of parallelism

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-07-03T21:38:58.118680Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-20T06:33:59.587034+00:00.