Paper Citation Record · LEDGER

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining

As of 23 August 2026, this Paper Citation Record lists 31 of 31 outbound references and 0 inbound Pith citation observations for arXiv:2607.07494.

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

pith.paper-citation-record.v1
2607.07494 v1

Coverage vector

measured 31 of 31 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-07-09T09:15:12.214083Z

measured 31 of 31 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-22T06:32:14.747728+00:00

measured 0 of 0 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links

measured 0 of 1 external citation measurements

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

Source: cited_works

Reference resolution

31 of 31 outbound references displayed

verified exact18
verified fuzzy5
unresolved0
parse uncertain0
malformed identifier2
metadata mismatch6

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation f2059980-4aad-474d-9191-13547e0f19a3 · outbound

This paper cites Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis

Reference 1

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.522575Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 25c15d21-ac6a-4772-999b-aa7130eb0aa7 · outbound

This paper cites Bandwidth optimal all-reduce algorithms for clusters of workstations,.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Bandwidth optimal all-reduce algorithms for clusters of workstations,

Reference 2

Resolution: verified exact; doi, observed 2026-07-09T09:16:06.530439Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation d0b93274-9f9b-475e-b361-3f92cbd9b5b7 · outbound

This paper cites InProceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2(Vancouver, BC, Canada)(ASPLOS 2023).

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining InProceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2(Vancouver, BC, Canada)(ASPLOS 2023)

Reference 3

Resolution: metadata mismatch; arxiv_id, observed 2026-07-09T09:16:06.475475Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 0a318c2e-7a83-4f3b-8fbf-d728b7ed2f33 · outbound

This paper cites FP8-LM: Training FP8 Large Language Models.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining FP8-LM: Training FP8 Large Language Models

Reference 4

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.500582Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation ddf31843-2c99-487e-b4f2-f2f4ac1398c0 · outbound

This paper cites COAT: Compressing Optimizer states and Activation for Memory-Efficient FP8 Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining COAT: Compressing Optimizer states and Activation for Memory-Efficient FP8 Training

Reference 5

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.523377Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 6c557088-e4b3-49a4-bcc5-9c1618a6957d · outbound

This paper cites Decoupled Weight Decay Regularization,.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Decoupled Weight Decay Regularization,

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-07-09T09:16:06.711251Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation c8ba83ef-4f75-4bbc-bd4e-b3b4c2d95e2a · outbound

This paper cites Decoupled Weight Decay Regularization.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Decoupled Weight Decay Regularization

Reference 7

Resolution: metadata mismatch; local_arxiv, observed 2026-07-09T09:16:06.503181Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 14715872-799e-4a17-9d7f-200112564f06 · outbound

This paper cites SDP4Bit: Toward 4-bit Communication Quantization in Sharded Data Parallelism for LLM Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining SDP4Bit: Toward 4-bit Communication Quantization in Sharded Data Parallelism for LLM Training

Reference 8

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.535859Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 93309472-6084-4e32-bf80-1856cef73652 · outbound

This paper cites Muon: An optimizer for hidden layers in neural networks,.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Muon: An optimizer for hidden layers in neural networks,

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-07-09T09:16:06.715190Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 028df0d8-96d5-43b3-95c1-2f594ef0e20b · outbound

This paper cites Available: https://kellerjordan.github.io/posts/muon/.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Available: https://kellerjordan.github.io/posts/muon/

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-07-09T09:16:06.712955Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 63e731fd-9438-4aac-b9ac-457c38815966 · outbound

This paper cites Amari,Information Geometry and Its Applications.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Amari,Information Geometry and Its Applications

Reference 11

Resolution: malformed identifier; raw_fallback, observed 2026-07-09T09:16:06.717572Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 5fe4aa30-baa6-4314-b6f0-26c35db697a5 · outbound

This paper cites Natural Gradient Works Efficiently in Learning , journal =.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Natural Gradient Works Efficiently in Learning , journal =

Reference 12

Resolution: verified exact; doi, observed 2026-07-09T09:16:06.528219Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T08:38:13.253949+00:00.

Observation 8782c345-28f4-4a7e-9dd9-83fdb7b5ad87 · outbound

This paper cites Optimizing Neural Networks with Kronecker-factored Approximate Curvature.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Optimizing Neural Networks with Kronecker-factored Approximate Curvature

Reference 13

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.515307Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 7ed52c58-f009-466d-870c-16cd34446578 · outbound

This paper cites Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM

Reference 14

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.525993Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation f9b0c093-bb8a-4dfe-8673-35189ed11d3e · outbound

This paper cites Openwebtext cor- pus,.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Openwebtext cor- pus,

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-07-09T09:16:06.705423Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 77bd7475-7b24-454f-81e8-3cc2d5443949 · outbound

This paper cites Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism

Reference 16

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.679697Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 5cf6404c-142b-481a-8678-195109daa2e2 · outbound

This paper cites Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism

Reference 17

Resolution: metadata mismatch; local_arxiv, observed 2026-07-09T09:16:06.482358Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation e4cca9e8-6a67-4820-8d2b-b420695220e3 · outbound

This paper cites GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism

Reference 18

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.520887Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 49378994-0105-4fc3-be1b-69c070a6aebb · outbound

This paper cites Memory-Efficient Pipeline-Parallel DNN Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Memory-Efficient Pipeline-Parallel DNN Training

Reference 19

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.540962Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 8ede1d75-5111-4ccb-bd94-561dca7c8fda · outbound

This paper cites ZeRO: Memory Optimizations Toward Training Trillion Parameter Models.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining ZeRO: Memory Optimizations Toward Training Trillion Parameter Models

Reference 20

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.530633Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation f72b09a6-2543-4ecb-aea5-a3711a6cba83 · outbound

This paper cites ZeRO++: Extremely Efficient Collective Communication for Giant Model Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining ZeRO++: Extremely Efficient Collective Communication for Giant Model Training

Reference 21

Resolution: malformed identifier; local_arxiv, observed 2026-07-09T09:16:06.676458Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation c31c9677-d310-4239-ab2b-88cf985e9f0a · outbound

This paper cites Kronecker-Factored Approximate Curvature for Modern Neural Network Architectures.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Kronecker-Factored Approximate Curvature for Modern Neural Network Architectures

Reference 22

Resolution: metadata mismatch; local_arxiv, observed 2026-07-09T09:16:06.535810Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation fde0184c-24a9-4b70-8164-6af20c5072b6 · outbound

This paper cites QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding

Reference 23

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.525204Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 384737c0-8c4b-4c26-b2e0-cc3911f32a6e · outbound

This paper cites Deep gradient compression: Reducing the communication bandwidth for distributed training,.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Deep gradient compression: Reducing the communication bandwidth for distributed training,

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-07-09T09:16:06.708056Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation cd1c83f9-abf0-49e7-af9b-d7af0fb478c3 · outbound

This paper cites Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Deep Gradient Compression: Reducing the Communication Bandwidth for Distributed Training

Reference 25

Resolution: metadata mismatch; local_arxiv, observed 2026-07-09T09:16:06.510183Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b8e742f2-10f4-40be-bcfc-b4d96c844f24 · outbound

This paper cites PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization

Reference 26

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.518204Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 4c5ab36d-92f2-4a45-9c18-3f0e2ae59ea7 · outbound

This paper cites 1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining 1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed

Reference 27

Resolution: metadata mismatch; local_arxiv, observed 2026-07-09T09:16:06.538354Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation b7609279-ffb8-41e6-87ae-ae7f4e5522a6 · outbound

This paper cites FP8 Formats for Deep Learning.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining FP8 Formats for Deep Learning

Reference 28

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.492807Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 0223999c-3059-413c-9253-5e6703b9c96a · outbound

This paper cites LoCo: Low-Bit Communication Adaptor for Large-scale Model Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining LoCo: Low-Bit Communication Adaptor for Large-scale Model Training

Reference 29

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.532939Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 75ae97cc-eea2-4299-99a6-6ba24ebfe5e2 · outbound

This paper cites Mixed Precision Training.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Mixed Precision Training

Reference 30

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.471623Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Observation 58e148da-2af6-44f8-aa17-242246c20812 · outbound

This paper cites Training and inference of large language models using 8-bit floating point.

GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining Training and inference of large language models using 8-bit floating point

Reference 31

Resolution: verified exact; local_arxiv, observed 2026-07-09T09:16:06.538522Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.

Pith citing papers

No inbound Pith citation observations are available.