Paper Citation Record · LEDGER

Heterogeneous Parallelism for Multimodal Large Language Model Training

As of 5 August 2026, this Paper Citation Record lists 23 of 23 outbound references and 0 inbound Pith citation observations for arXiv:2605.27678.

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

pith.paper-citation-record.v1
2605.27678 v1

Coverage vector

measured 23 of 23 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-06-29T18:42:09.591282Z

measured 23 of 23 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-05T06:32:48.257954+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

23 of 23 outbound references displayed

verified exact12
verified fuzzy0
unresolved8
parse uncertain0
malformed identifier1
metadata mismatch2

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 37afc617-2f4a-4cc3-b942-e946c76673ae · outbound

This paper cites Chiang, Z.

Heterogeneous Parallelism for Multimodal Large Language Model Training Chiang, Z

Reference 1

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b466a14f-5a54-4f85-a123-fc09a5a382e7 · outbound

This paper cites Using Cornstarch 5d parallelism.https://cornstarch-org.github.io/ parallelization/cornstarch_parallel/, 2025.

Heterogeneous Parallelism for Multimodal Large Language Model Training Using Cornstarch 5d parallelism.https://cornstarch-org.github.io/ parallelization/cornstarch_parallel/, 2025

Reference 2

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 835a55c5-6677-4da3-8d02-6c760d47a817 · outbound

This paper cites an unresolved cited work.

Heterogeneous Parallelism for Multimodal Large Language Model Training Unresolved cited work

Reference 3

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dbd5eeaf-f463-4c91-a67a-0f1e61aace7b · outbound

This paper cites Huang, Z.

Heterogeneous Parallelism for Multimodal Large Language Model Training Huang, Z

Reference 4

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bd706786-a3c5-4442-9be2-0de0d60f110f · outbound

This paper cites an unresolved cited work.

Heterogeneous Parallelism for Multimodal Large Language Model Training Unresolved cited work

Reference 5

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 602a3a2e-ac59-43c7-a30e-31c746f92d97 · outbound

This paper cites Efficient Distributed MLLM Training with Cornstarch.

Heterogeneous Parallelism for Multimodal Large Language Model Training Efficient Distributed MLLM Training with Cornstarch

Reference 6

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.404654Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 5fe3c342-442d-4adf-9c38-cc2c6e1448e5 · outbound

This paper cites GraphPipe: Improving Performance and Scalability of DNN Training with Graph Pipeline Parallelism.

Heterogeneous Parallelism for Multimodal Large Language Model Training GraphPipe: Improving Performance and Scalability of DNN Training with Graph Pipeline Parallelism

Reference 7

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.399012Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 49192dec-bb94-41df-af56-6e74e88f3818 · outbound

This paper cites DCP: Addressing Input Dynamism In Long-Context Training via Dynamic Context Parallelism , url=.

Heterogeneous Parallelism for Multimodal Large Language Model Training DCP: Addressing Input Dynamism In Long-Context Training via Dynamic Context Parallelism , url=

Reference 8

Resolution: metadata mismatch; arxiv_id, observed 2026-06-29T18:43:50.190743Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation a0a73dd6-7e11-44b1-bf68-2f434cd09e32 · outbound

This paper cites an unresolved cited work.

Heterogeneous Parallelism for Multimodal Large Language Model Training Unresolved cited work

Reference 9

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.414857Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 18dd1c7d-1d54-4fb3-b41d-d0ad1f22525a · outbound

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

Heterogeneous Parallelism for Multimodal Large Language Model Training Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM

Reference 10

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.412459Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 9c284704-0d17-4296-b8e5-f12cb6ab10f6 · outbound

This paper cites Efficient Scaling of LLM Training with Flexible Context Parallelism.

Heterogeneous Parallelism for Multimodal Large Language Model Training Efficient Scaling of LLM Training with Flexible Context Parallelism

Reference 11

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.408967Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 016ac0a1-b5d2-49a4-bcd4-eeb4077c09ee · outbound

This paper cites Radford, J.

Heterogeneous Parallelism for Multimodal Large Language Model Training Radford, J

Reference 12

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation eabf0c20-599f-4976-baae-a1fa4f4e07f8 · outbound

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

Heterogeneous Parallelism for Multimodal Large Language Model Training ZeRO: Memory Optimizations Toward Training Trillion Parameter Models

Reference 13

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.409788Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation f328d04f-1a59-481b-87cc-bf24da721425 · outbound

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

Heterogeneous Parallelism for Multimodal Large Language Model Training Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism

Reference 14

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.417019Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 16bc8f80-e4bd-46ec-bb85-73c17165dfd3 · outbound

This paper cites Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model.

Heterogeneous Parallelism for Multimodal Large Language Model Training Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model

Reference 15

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.407336Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 8ae92075-ed80-43b8-944a-6720f4976e71 · outbound

This paper cites FlexSP: Accelerating Large Language Model Training via Flexible Sequence Parallelism.

Heterogeneous Parallelism for Multimodal Large Language Model Training FlexSP: Accelerating Large Language Model Training via Flexible Sequence Parallelism

Reference 16

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.391035Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 7b10cdd8-5ac1-4854-99cb-247968f4ab06 · outbound

This paper cites an unresolved cited work.

Heterogeneous Parallelism for Multimodal Large Language Model Training Unresolved cited work

Reference 17

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6f023501-bd6c-474c-b4ff-e2ec066ff998 · outbound

This paper cites an unresolved cited work.

Heterogeneous Parallelism for Multimodal Large Language Model Training Unresolved cited work

Reference 18

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.396272Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 15516d98-1865-4fab-a2ed-cd8b078188a6 · outbound

This paper cites Dfvg: A heterogeneous architecture for speculative decoding with draft-on-fpga and verify-on-gpu,.

Heterogeneous Parallelism for Multimodal Large Language Model Training Dfvg: A heterogeneous architecture for speculative decoding with draft-on-fpga and verify-on-gpu,

Reference 19

Resolution: metadata mismatch; arxiv_id, observed 2026-06-29T18:43:50.190089Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 59ec43dd-82bd-4913-bf9b-5cbabce0c2c1 · outbound

This paper cites Disttrain: Addressing model and data heterogeneity with disaggregated training for multimodal large language models.

Heterogeneous Parallelism for Multimodal Large Language Model Training Disttrain: Addressing model and data heterogeneity with disaggregated training for multimodal large language models

Reference 20

Resolution: verified exact; arxiv_id, observed 2026-06-29T18:43:50.411473Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation f7e61979-09bc-49e5-929c-4ecec8b13f75 · outbound

This paper cites MegaScale-Data: Scaling Dataloader for Multisource Large Foundation Model Training.

Heterogeneous Parallelism for Multimodal Large Language Model Training MegaScale-Data: Scaling Dataloader for Multisource Large Foundation Model Training

Reference 21

Resolution: verified exact; local_arxiv, observed 2026-06-29T18:43:50.417700Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation 63d514d6-95ec-44e9-aff1-b57c1caed6b8 · outbound

This paper cites PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel.

Heterogeneous Parallelism for Multimodal Large Language Model Training PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel

Reference 22

Resolution: malformed identifier; local_arxiv, observed 2026-06-29T18:43:50.403974Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-05T06:32:48.257954+00:00.

Observation c3758176-ef30-45cb-8fdc-d6c5bffa8d47 · outbound

This paper cites language.

Heterogeneous Parallelism for Multimodal Large Language Model Training language

Reference 23

Resolution: unresolved; no resolver link, observed 2026-06-29T18:42:09.591282Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

No inbound Pith citation observations are available.