SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam | Paper Citation Record

Observation e944ecc4-1240-4dbb-8a4b-37159c5297a8 · outbound

This paper cites Adam: A method for stochastic optimization,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adam: A method for stochastic optimization,

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.201901Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e6e02f0c-d61c-4d6f-80e1-ea04028cd716 · outbound

This paper cites Attention is all you need,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Attention is all you need,

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.191786Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 130b3ed3-5f82-4fe6-8a9f-e2350eb1faee · outbound

This paper cites Language mod- els are few-shot learners,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Language mod- els are few-shot learners,

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.458605Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1a5ec753-fa43-4e58-b12d-202711d646bf · outbound

This paper cites Palm: Scal- ing language modeling with pathways,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Palm: Scal- ing language modeling with pathways,

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.462151Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7091e28a-58fc-431a-b9f6-edc9e96e4b74 · outbound

This paper cites The Llama 3 Herd of Models.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam The Llama 3 Herd of Models

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.465335Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a7f18002-e37c-421a-998c-93439602f0dd · outbound

This paper cites DeepSeek-V3 Technical Report.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam DeepSeek-V3 Technical Report

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.468721Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c344c854-4c34-4e60-8ee5-4d7fa5be9619 · outbound

This paper cites Learning transferable visual models from natural language supervision,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Learning transferable visual models from natural language supervision,

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.471944Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 99a15b1d-518d-42bd-9632-e3469bc87f4b · outbound

This paper cites Segment Anything.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Segment Anything

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.474845Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c5b1c58-700a-4826-b81c-372942e56358 · outbound

This paper cites A convnet for the 2020s,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A convnet for the 2020s,

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.478341Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ea6ca0d-f1a1-4b7c-acfd-602228e86852 · outbound

This paper cites Convnext v2: Co-designing and scaling convnets with masked autoen- coders,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Convnext v2: Co-designing and scaling convnets with masked autoen- coders,

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.481830Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c1825759-aff8-4476-ad47-2cf31d09d429 · outbound

This paper cites Imagenet classification with deep convolutional neural networks,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Imagenet classification with deep convolutional neural networks,

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.485548Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c49b5738-5629-45bf-9d39-f9530cbe0292 · outbound

This paper cites Deep residual learning for image recognition,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Deep residual learning for image recognition,

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.488397Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 95fdae10-305a-4e29-92b0-50fbe1e9c141 · outbound

This paper cites Symbolic Discovery of Optimization Algorithms.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Symbolic Discovery of Optimization Algorithms

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.490853Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 387290f7-6e61-4fbb-9549-3f7e6b456b20 · outbound

This paper cites Noise Is Not the Main Factor Behind the Gap Between SGD and Adam on Transformers, but Sign Descent Might Be.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Noise Is Not the Main Factor Behind the Gap Between SGD and Adam on Transformers, but Sign Descent Might Be

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.493852Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 64c549bf-2a0d-45a2-921e-0411c3fa38b6 · outbound

This paper cites Adaptive subgradient methods for online learning and stochastic optimization.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adaptive subgradient methods for online learning and stochastic optimization

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.497361Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9034aef6-356b-4b25-bf4d-b3a7e75faeb4 · outbound

This paper cites Neural networks for machine learning lecture 6a overview of mini-batch gradient descent,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Neural networks for machine learning lecture 6a overview of mini-batch gradient descent,

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.137025Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 8ba9cd53-5f3c-4235-a11e-ac7c73ec22b0 · outbound

This paper cites ADADELTA: An Adaptive Learning Rate Method.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam ADADELTA: An Adaptive Learning Rate Method

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.502866Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8fad584c-c762-4759-acd9-db97b754c220 · outbound

This paper cites Incorporating nesterov momentum into adam,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Incorporating nesterov momentum into adam,

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.124513Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation b424d17a-bdd7-4d6d-84f5-938ee379f839 · outbound

This paper cites On the convergence of adam and beyond,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam On the convergence of adam and beyond,

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.509896Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6bf20935-3052-4582-a149-eba073699fa4 · outbound

This paper cites Decoupled Weight Decay Regularization.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Decoupled Weight Decay Regularization

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.513195Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 882ac5b3-d99a-4380-afc0-42040410dad4 · outbound

This paper cites Adabelief optimizer: Adapting stepsizes by the belief in observed gradients,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adabelief optimizer: Adapting stepsizes by the belief in observed gradients,

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.109013Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation f6cc7300-be44-4848-8327-9827f0c8109d · outbound

This paper cites Adafactor: Adaptive learning rates with sub- linear memory cost,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adafactor: Adaptive learning rates with sub- linear memory cost,

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.097992Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation d39444e8-0734-4f72-98ad-07c53c256862 · outbound

This paper cites 1-bit stochastic gradient descent and its application to data-parallel distributed training of speech dnns,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam 1-bit stochastic gradient descent and its application to data-parallel distributed training of speech dnns,

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.088592Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation ebea6fa2-4ce9-47dd-be8e-5ba659eea44f · outbound

This paper cites Signsgd: Compressed optimisation for non-convex problems,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Signsgd: Compressed optimisation for non-convex problems,

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.078021Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation e926c48a-d74f-4e96-8149-97437b6e9db3 · outbound

This paper cites Momentum ensures convergence of SIGNSGD under weaker assumptions,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Momentum ensures convergence of SIGNSGD under weaker assumptions,

Reference 25

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.067344Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 8a4e6230-10b8-41d4-9d9c-120d5a4c958b · outbound

This paper cites Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.530074Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 82b4b94f-b472-4f8f-addb-70900f8590bd · outbound

This paper cites A direct adaptive method for faster backpropagation learning: The rprop algorithm,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A direct adaptive method for faster backpropagation learning: The rprop algorithm,

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.055958Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 8824a63e-45a6-459f-97fd-1ec59887e172 · outbound

This paper cites 1-bit stochastic gradient descent and its application to data-parallel distributed training of speech DNNs.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam 1-bit stochastic gradient descent and its application to data-parallel distributed training of speech DNNs

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.045610Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation afb2ed8b-b8f9-47ed-bb3d-4140897acc52 · outbound

This paper cites Lion Secretly Solves Constrained Optimization: As Lyapunov Predicts.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Lion Secretly Solves Constrained Optimization: As Lyapunov Predicts

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.538230Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6036972d-a558-4688-963c-b919c3331931 · outbound

This paper cites Dissecting Adam: The sign, magnitude and variance of stochastic gradients,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Dissecting Adam: The sign, magnitude and variance of stochastic gradients,

Reference 30

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.035995Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 03e8a653-b280-4c7a-9d79-158e223e8596 · outbound

This paper cites Heavy-Tailed Class Imbalance and Why Adam Outperforms Gradient Descent on Language Models.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Heavy-Tailed Class Imbalance and Why Adam Outperforms Gradient Descent on Language Models

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.543726Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0b7b1394-3226-48cb-a8e0-b8b9b09a1f9a · outbound

This paper cites A method of solving a convex programming problem with convergence rate o\bigl(kˆ2\bigr),.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A method of solving a convex programming problem with convergence rate o\bigl(kˆ2\bigr),

Reference 32

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.025018Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 9d6fc65c-1ed3-4412-add0-8e36cd225a20 · outbound

This paper cites Springer Science & Business Media, 2013, vol.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Springer Science & Business Media, 2013, vol

Reference 33

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.015378Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 1f9894cf-5e9c-4c60-9bb3-92b9b905aecf · outbound

This paper cites Adan: Adaptive nesterov momentum algorithm for faster optimizing deep models,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adan: Adaptive nesterov momentum algorithm for faster optimizing deep models,

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:17.003645Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 17c3485f-f796-4d31-97c8-1e15f18f969d · outbound

This paper cites Win: Weight-decay-integrated nesterov acceleration for adaptive gradient algorithms,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Win: Weight-decay-integrated nesterov acceleration for adaptive gradient algorithms,

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.991917Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 7e6bf4ef-121e-415e-a5f8-4efda35ca0fd · outbound

This paper cites GLM-130B: An Open Bilingual Pre-trained Model.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam GLM-130B: An Open Bilingual Pre-trained Model

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.558707Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fef1eb67-c197-4ee5-84c5-067f88733fc4 · outbound

This paper cites LLaMA: Open and Efficient Foundation Language Models.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam LLaMA: Open and Efficient Foundation Language Models

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.561798Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4389efc2-d313-4583-8363-821f45cddab5 · outbound

This paper cites Baichuan 2: Open Large-scale Language Models.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Baichuan 2: Open Large-scale Language Models

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.564494Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a65fb9e5-510a-45de-9c7b-e09761f73eb6 · outbound

This paper cites What Language Model to Train if You Have One Million GPU Hours?.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam What Language Model to Train if You Have One Million GPU Hours?

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.567629Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6c436ba2-b562-42f2-af4e-208e2df607d3 · outbound

This paper cites A Theory on Adam Instability in Large-Scale Machine Learning.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A Theory on Adam Instability in Large-Scale Machine Learning

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.570768Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4a9a43cc-6d72-4ab3-a56a-d57d4b7aaa05 · outbound

This paper cites A Mean Field Theory of Batch Normalization.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A Mean Field Theory of Batch Normalization

Reference 41

Resolution: verified exact; local_arxiv, observed 2026-08-06T19:10:16.753519Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 81412f7d-d3c9-443f-b776-78d06309f9d7 · outbound

This paper cites Understanding the Difficulty of Training Transformers.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Understanding the Difficulty of Training Transformers

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.577022Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ae162317-639b-4c10-8aa0-87babd01ece8 · outbound

This paper cites On layer normalization in the trans- former architecture,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam On layer normalization in the trans- former architecture,

Reference 43

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.981559Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 2652fda8-ca7c-4f19-9ce7-dd05b222d562 · outbound

This paper cites The lipschitz constant of self- attention,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam The lipschitz constant of self- attention,

Reference 44

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.969465Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 8dfbbdde-7c7e-47bd-8fd7-2af0221cade8 · outbound

This paper cites LipsFormer: Introducing Lipschitz Continuity to Vision Transformers.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam LipsFormer: Introducing Lipschitz Continuity to Vision Transformers

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.586759Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8c0381e7-1410-476e-9265-b2dbab2a4363 · outbound

This paper cites Signal propagation in transformers: Theoretical perspectives and the role of rank collapse,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Signal propagation in transformers: Theoretical perspectives and the role of rank collapse,

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.955532Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 218ef676-cdc5-4fff-9afb-a88f3cad0e79 · outbound

This paper cites On the Variance of the Adaptive Learning Rate and Beyond.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam On the Variance of the Adaptive Learning Rate and Beyond

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.591929Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b1702c3e-1070-4189-83a0-ceba8119b40e · outbound

This paper cites Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.595651Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ee2e7d3e-cdc5-4a94-90b3-ffc1a0050369 · outbound

This paper cites Loss Spike in Training Neural Networks.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Loss Spike in Training Neural Networks

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.598782Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 05f8d309-aa58-4cf9-ab76-89e4d31a7e70 · outbound

This paper cites Stochasticity of deterministic gradient descent: Large learning rate for multiscale objective function,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Stochasticity of deterministic gradient descent: Large learning rate for multiscale objective function,

Reference 50

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.945621Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 0665f40e-517a-40fa-9b6b-57ba587a69f1 · outbound

This paper cites On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.605575Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation aa8ccb70-02cc-4385-9382-1b1296c7d6a0 · outbound

This paper cites A Simple Convergence Proof of Adam and Adagrad.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A Simple Convergence Proof of Adam and Adagrad

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.608689Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2da32d6b-006e-4c34-9b9f-d5b321245e07 · outbound

This paper cites Convergence guarantees for RMSProp and ADAM in non-convex optimization and an empirical comparison to Nesterov acceleration.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Convergence guarantees for RMSProp and ADAM in non-convex optimization and an empirical comparison to Nesterov acceleration

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.611740Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c793896d-c0f0-4540-b1f4-ba65969027a9 · outbound

This paper cites Adam can converge without any modification on update rules,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Adam can converge without any modification on update rules,

Reference 54

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.935092Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation edb696f8-338f-4603-bc9b-65aa87994a2a · outbound

This paper cites Convergence of adam under relaxed assumptions,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Convergence of adam under relaxed assumptions,

Reference 55

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.925114Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation aef5e537-b67a-4191-baf8-44b8ee68e314 · outbound

This paper cites On Convergence of Adam for Stochastic Optimization under Relaxed Assumptions.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam On Convergence of Adam for Stochastic Optimization under Relaxed Assumptions

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.619880Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d412bbce-b75e-4338-85a1-8d95ec0c4527 · outbound

This paper cites Lower bounds for non-convex stochastic optimization,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Lower bounds for non-convex stochastic optimization,

Reference 57

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.915141Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Observation 8aa29741-fb40-45e8-9b6f-6eb49c4d653a · outbound

This paper cites A stochastic approximation method,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam A stochastic approximation method,

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.625262Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 37318d86-763a-4520-a46e-43cf4fa59af7 · outbound

This paper cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-06T19:10:16.628081Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ef273616-c7fa-4b26-957c-cf3618c69d17 · outbound

This paper cites Language models are unsupervised multitask learners,.

SoftSignSGD(S3): An Enhanced Optimizer for Practical DNN Training and Loss Spikes Minimization Beyond Adam Language models are unsupervised multitask learners,

Reference 60

Resolution: verified fuzzy; raw_fallback, observed 2026-08-06T19:10:16.898796Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.

Coverage vector

Reference resolution

Outbound references

Pith citing papers