Paper Citation Record · LEDGER

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation

As of 19 August 2026, this Paper Citation Record lists 55 of 55 outbound references and 3 inbound Pith citation observations for arXiv:2504.19232.

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

pith.paper-citation-record.v1
2504.19232 v1

Coverage vector

measured 55 of 55 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-16T06:04:35.588702Z

measured 58 of 58 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-19T06:32:44.657259+00:00

measured 3 of 3 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-05-12T04:01:42.265205Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-05-12T06:41:44.858043Z

Reference resolution

55 of 55 outbound references displayed

verified exact0
verified fuzzy26
unresolved28
parse uncertain1
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation d74d4bdf-0c4b-40ab-a904-dba4f3ec18a3 · outbound

This paper cites GPT-4 Technical Report.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation GPT-4 Technical Report

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.351955Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cd9bf8e2-e655-427f-a258-10a7ea3419f6 · outbound

This paper cites Ethereal: Divide and Conquer Network Load Balancing in Large-Scale Distributed Training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Ethereal: Divide and Conquer Network Load Balancing in Large-Scale Distributed Training

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.357100Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 45a9577d-22e3-4bb9-8dc3-cd57c172b3f0 · outbound

This paper cites Conga: Distributed congestion-aware load balancing for datacenters.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Conga: Distributed congestion-aware load balancing for datacenters

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.545686Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 1b68bad4-bba1-4888-8417-30458ed933ea · outbound

This paper cites Varuna: scal- able, low-cost training of massive deep learning models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Varuna: scal- able, low-cost training of massive deep learning models

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.366251Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7809112a-468a-41f7-999c-f33166243c4d · outbound

This paper cites Crux: Gpu-efficient communication scheduling for deep learning training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Crux: Gpu-efficient communication scheduling for deep learning training

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.519218Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 44d2fda3-0bc3-48b7-b1b1-b0ee2880c92e · outbound

This paper cites Revisiting Distributed Synchronous SGD.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Revisiting Distributed Synchronous SGD

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.375253Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e2c35aa4-f977-4e31-8e7e-e0ec3368dff0 · outbound

This paper cites Mscclang: Microsoft collective communication language.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Mscclang: Microsoft collective communication language

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.380526Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 973776e0-62b1-4b68-8a01-d8f9ea148ff6 · outbound

This paper cites Xputimer: Anomaly diagnostics for divergent llm training in gpu clusters of thousand-plus scale.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Xputimer: Anomaly diagnostics for divergent llm training in gpu clusters of thousand-plus scale

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.384723Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d6756975-bfa0-402f-989c-0805f5af2d65 · outbound

This paper cites On the impact of packet spraying in data center networks.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation On the impact of packet spraying in data center networks

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.493711Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation e536f1e9-d560-4c82-a803-03e1b91c912c · outbound

This paper cites ACCL: Architecting highly scalable distributed training systems with highly efficient collective communication library.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation ACCL: Architecting highly scalable distributed training systems with highly efficient collective communication library

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.477648Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 7d9c6fce-38ec-4ed8-b4ae-d9d48b7a4d0b · outbound

This paper cites The Llama 3 Herd of Models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation The Llama 3 Herd of Models

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.397576Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bcf63b55-1d15-42e7-8a11-7a182ad336cf · outbound

This paper cites Recycle: Resilient training of large dnns using pipeline adaptation.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Recycle: Resilient training of large dnns using pipeline adaptation

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.460550Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 44952ce9-bdd7-4543-9b6a-1f601503e472 · outbound

This paper cites Rdma over ethernet for distributed training at meta scale.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Rdma over ethernet for distributed training at meta scale

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.444051Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 73d7c4e9-bb42-4978-8d66-4dbf16cab18f · outbound

This paper cites DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.410895Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cf75e0af-1757-44b2-86f7-417c793ba105 · outbound

This paper cites Addressing the straggler problem for iterative convergent parallel ml.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Addressing the straggler problem for iterative convergent parallel ml

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.427808Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation ee3a00ec-ed5b-496f-8ef4-de871c17d87d · outbound

This paper cites Characterization of large language model development in the datacenter.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Characterization of large language model development in the datacenter

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.412562Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 112058d7-43e0-4434-96ac-2bbe43c99399 · outbound

This paper cites Gpipe: Effi- cient training of giant neural networks using pipeline parallelism.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Gpipe: Effi- cient training of giant neural networks using pipeline parallelism

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.423164Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6355f156-879e-4286-9c6a-3e9a86cffaaf · outbound

This paper cites Gloo, 2025.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Gloo, 2025

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.388430Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation f8f8e873-a8dd-4d8b-8bda-31c6d6a239d2 · outbound

This paper cites Oobleck: Resilient distributed training of large models using pipeline templates.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Oobleck: Resilient distributed training of large models using pipeline templates

Reference 19

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.372362Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 9702dcae-e710-4eb5-aecf-5d957ea62a85 · outbound

This paper cites Analysis of{Large-Scale}{Multi-Tenant}{GPU} clus- ters for{DNN} training workloads.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Analysis of{Large-Scale}{Multi-Tenant}{GPU} clus- ters for{DNN} training workloads

Reference 20

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.357838Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation ecd09150-628d-4146-bc81-255a900b04a9 · outbound

This paper cites MegaScale: Scaling large language model training to more than 10,000{GPUs}.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation MegaScale: Scaling large language model training to more than 10,000{GPUs}

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.342295Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 71b0085e-0de1-49fb-a5c5-eb433a1324b3 · outbound

This paper cites Scaling Laws for Neural Language Models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Scaling Laws for Neural Language Models

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.444019Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7a3cd834-edb9-4103-af28-232065c24e4f · outbound

This paper cites STrack: A Reliable Multipath Transport for AI/ML Clusters.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation STrack: A Reliable Multipath Transport for AI/ML Clusters

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.448686Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 082547a8-f895-45c3-9a47-7fe1f1ec5f78 · outbound

This paper cites Malleus: Straggler-Resilient Hybrid Parallel Training of Large-scale Models via Malleable Data and Model Parallelization.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Malleus: Straggler-Resilient Hybrid Parallel Training of Large-scale Models via Malleable Data and Model Parallelization

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.453139Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9cb8ee2c-6895-4b44-b563-c18723bbc92f · outbound

This paper cites TorchTitan: One-stop PyTorch native solution for production ready LLM pre-training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation TorchTitan: One-stop PyTorch native solution for production ready LLM pre-training

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.457714Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f308572a-6deb-490e-8966-4122a000f855 · outbound

This paper cites DeepSeek-V3 Technical Report.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation DeepSeek-V3 Technical Report

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.462080Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b9767681-5a80-403a-b1fb-1ec3739013fb · outbound

This paper cites Sdpipe: A semi-decentralized framework for heterogeneity-aware pipeline-parallel training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Sdpipe: A semi-decentralized framework for heterogeneity-aware pipeline-parallel training

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.328178Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation a1de021c-919a-4da1-bff1-a1e5f6af30db · outbound

This paper cites {CheckFreq}: Frequent, {Fine- Grained}{DNN} checkpointing.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation {CheckFreq}: Frequent, {Fine- Grained}{DNN} checkpointing

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.313632Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 7a7e9161-6c13-40ba-961c-fa24856e6798 · outbound

This paper cites Pipedream: Gen- eralized pipeline parallelism for dnn training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Pipedream: Gen- eralized pipeline parallelism for dnn training

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.474543Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6d65604b-10e8-4a69-9a6f-affcf60a9ecd · outbound

This paper cites Efficient large-scale language model training on gpu clusters using megatron-lm.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Efficient large-scale language model training on gpu clusters using megatron-lm

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.478525Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d59b5674-c390-4f1a-9c63-1febb4337fb7 · outbound

This paper cites Nvidia nsight systems, 2025.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Nvidia nsight systems, 2025

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.278860Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation b2b7cb45-6cc8-41e2-ad3f-544024a9804b · outbound

This paper cites Openai sora, 2024.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Openai sora, 2024

Reference 32

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.263552Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 06708ac2-2204-45ea-ade2-2209f032de1c · outbound

This paper cites Zero Bubble Pipeline Parallelism.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Zero Bubble Pipeline Parallelism

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.491430Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9d71e85f-cffc-4bc4-be41-8b08ea2abc3c · outbound

This paper cites Alibaba hpn: A data center network for large language model training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Alibaba hpn: A data center network for large language model training

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.496340Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a2e8750e-f5d9-4e11-b680-3bad814175d0 · outbound

This paper cites In 21st USENIX Sympo- sium on Networked Systems Design and Implementation (NSDI 24), pages 1403–1420, 2024.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation In 21st USENIX Sympo- sium on Networked Systems Design and Implementation (NSDI 24), pages 1403–1420, 2024

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.239147Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 2f9c70b8-9d18-4b4e-826d-8e046b4b5d70 · outbound

This paper cites Zero: Memory optimizations toward train- ing trillion parameter models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Zero: Memory optimizations toward train- ing trillion parameter models

Reference 36

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.224364Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 8c1fef91-ae7e-4a07-b410-3b4acb8b9a57 · outbound

This paper cites Redis - the real-time data platform,.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Redis - the real-time data platform,

Reference 37

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.210074Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation fdae3e64-3c20-45fb-adfb-1abcb093ec01 · outbound

This paper cites BLOOM: A 176B-Parameter Open-Access Multilingual Language Model.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation BLOOM: A 176B-Parameter Open-Access Multilingual Language Model

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.518187Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2c12f68b-b558-49e5-bd2f-b6c6b7e8d598 · outbound

This paper cites 15 {TACCL}: Guiding collective algorithm synthesis us- ing communication sketches.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation 15 {TACCL}: Guiding collective algorithm synthesis us- ing communication sketches

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.180837Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 3636a8e1-ff83-4123-945e-32b278e264c9 · outbound

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

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.526609Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 008d5325-e487-4053-a835-10295e13b62d · outbound

This paper cites Effective multi-gpu communication using multiple cuda streams and threads.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Effective multi-gpu communication using multiple cuda streams and threads

Reference 41

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.165947Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 478520f2-4b7f-4eac-be7e-bdb029ffc66d · outbound

This paper cites Gemini: A Family of Highly Capable Multimodal Models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Gemini: A Family of Highly Capable Multimodal Models

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.535751Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 91639a0a-436e-4d98-bb9e-897f3127be7a · outbound

This paper cites Multipath issues in unicast and multicast next-hop selection.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Multipath issues in unicast and multicast next-hop selection

Reference 43

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.149049Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation abd4ffa9-2844-4527-aafd-908623d2144a · outbound

This paper cites Bamboo: Making preemptible in- stances resilient for affordable training of large{DNNs}.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Bamboo: Making preemptible in- stances resilient for affordable training of large{DNNs}

Reference 44

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.132596Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 737b7469-9cca-4a80-a317-cb61d54cfe71 · outbound

This paper cites an unresolved cited work.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Unresolved cited work

Reference 45

Resolution: unresolved; raw_fallback, observed 2026-08-16T06:04:36.116305Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation efec145c-f2d8-4a5b-9e57-5ce95568108f · outbound

This paper cites Machine Learning Model Sizes and the Parameter Gap.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Machine Learning Model Sizes and the Parameter Gap

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.553102Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 83dbe049-21b1-4a10-887e-0a75e01576b2 · outbound

This paper cites an unresolved cited work.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Unresolved cited work

Reference 47

Resolution: unresolved; raw_fallback, observed 2026-08-16T06:04:36.101833Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 515c1cfd-aeee-40a1-907d-199fd3edadb7 · outbound

This paper cites FALCON: Pinpointing and Mitigating Stragglers for Large-Scale Hybrid-Parallel Training.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation FALCON: Pinpointing and Mitigating Stragglers for Large-Scale Hybrid-Parallel Training

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.561757Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dd9a87e4-729d-4d9e-a12c-e8ec670a6cca · outbound

This paper cites SuperBench: Improving cloud AI infrastructure reliability with proactive valida- tion.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation SuperBench: Improving cloud AI infrastructure reliability with proactive valida- tion

Reference 49

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.086589Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 01713743-2cfc-4cd6-a2a1-d3f3ba13f9a0 · outbound

This paper cites OPT: Open Pre-trained Transformer Language Models.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation OPT: Open Pre-trained Transformer Language Models

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.570749Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a4661709-0b82-40d1-a79c-3b14c627054d · outbound

This paper cites Deepep: an efficient expert- parallel communication library.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Deepep: an efficient expert- parallel communication library

Reference 51

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.071765Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 76aceb5e-156f-4b05-b449-495db70c8516 · outbound

This paper cites Alpa: Automating inter-and{Intra-Operator} parallelism for distributed deep learning.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Alpa: Automating inter-and{Intra-Operator} parallelism for distributed deep learning

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.579912Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c6a1ef8f-7425-4c68-84b8-56fdcbf10192 · outbound

This paper cites Optimizing RLHF Training for Large Language Models with Stage Fusion.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Optimizing RLHF Training for Large Language Models with Stage Fusion

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-16T06:04:35.584122Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ac0ef10c-bb91-4194-bd6b-b677f151c632 · outbound

This paper cites Falcon: Addressing strag- glers in heterogeneous parameter server via multiple par- allelism.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Falcon: Addressing strag- glers in heterogeneous parameter server via multiple par- allelism

Reference 54

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T06:04:36.047577Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 180f88ec-4369-4ab4-af3f-02594c6bc224 · outbound

This paper cites an unresolved cited work.

Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation Unresolved cited work

Reference 2009

Resolution: parse uncertain; raw_fallback, observed 2026-08-16T06:04:36.194887Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Pith citing papers

Observation 993329e5-db85-4264-97e6-8a6aff9b9b9c · inbound

ResiHP: Taming LLM Training Failures with Dynamic Hybrid Parallelism cites this paper.

ResiHP: Taming LLM Training Failures with Dynamic Hybrid Parallelism Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation

Reference 49

Resolution: verified exact; arxiv_id, observed 2026-05-11T21:31:15.912603Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 6b6f1df8-4c35-4b50-8767-088c04899e1b · inbound

ResiHP: Taming LLM Training Failures with Dynamic Hybrid Parallelism cites this paper.

ResiHP: Taming LLM Training Failures with Dynamic Hybrid Parallelism Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation

Reference 49

Resolution: verified exact; arxiv_id, observed 2026-05-12T06:41:44.863146Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.

Observation 4fdc623d-d305-4449-aa6e-ea5e6b0026f3 · inbound

MegaScale-Omni: A Hyper-Scale, Workload-Resilient System for MultiModal LLM Training in Production cites this paper.

MegaScale-Omni: A Hyper-Scale, Workload-Resilient System for MultiModal LLM Training in Production Adaptra: Straggler-Resilient Hybrid-Parallel Training with Pipeline Adaptation

Reference 57

Resolution: verified exact; arxiv_id, observed 2026-05-12T02:06:14.954658Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.