Paper Citation Record · LEDGER

Why Do We Need Warm-up? A Theoretical Perspective

As of 7 August 2026, this Paper Citation Record lists 76 of 76 outbound references and 2 inbound Pith citation observations for arXiv:2510.03164.

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

pith.paper-citation-record.v1
2510.03164 v2

Coverage vector

measured 76 of 76 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-04T12:39:03.472216Z

measured 78 of 78 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-07T06:34:17.273281+00:00

measured 2 of 2 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-06-30T20:13:36.440752Z

measured 0 of 1 external citation measurements

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

Source: pith, observed 2026-06-30T20:15:04.536047Z

Reference resolution

76 of 76 outbound references displayed

verified exact0
verified fuzzy0
unresolved76
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 728eded1-397e-4d7f-8e45-3cc1e04f527b · outbound

This paper cites plainlm: Language model pretraining in pytorch.

Why Do We Need Warm-up? A Theoretical Perspective plainlm: Language model pretraining in pytorch

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:52.774743Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 62a51332-e1c6-4389-b70f-fc7e06fd687a · outbound

This paper cites vision: Vision model pretraining in pytorch.

Why Do We Need Warm-up? A Theoretical Perspective vision: Vision model pretraining in pytorch

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:52.939248Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 431fc2e4-8aa5-48a6-ad74-3cd69f9691da · outbound

This paper cites Benefits of learning rate annealing for tuning-robustness in stochastic optimization.

Why Do We Need Warm-up? A Theoretical Perspective Benefits of learning rate annealing for tuning-robustness in stochastic optimization

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.057902Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 149fb9d5-ab9c-4174-9e23-7fbde6b68020 · outbound

This paper cites Layer Normalization.

Why Do We Need Warm-up? A Theoretical Perspective Layer Normalization

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.264531Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a889a417-0937-40e6-a32c-1016652dc328 · outbound

This paper cites A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets.

Why Do We Need Warm-up? A Theoretical Perspective A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.411038Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e74688c1-4fe4-4611-9377-81ff0ab82c92 · outbound

This paper cites On the Interaction of Batch Noise, Adaptivity, and Compression, under $(L_0,L_1)$-Smoothness: An SDE Approach.

Why Do We Need Warm-up? A Theoretical Perspective On the Interaction of Batch Noise, Adaptivity, and Compression, under $(L_0,L_1)$-Smoothness: An SDE Approach

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.610360Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 57c806d0-96f4-4a92-99f6-aedbce155fcc · outbound

This paper cites Why Gradients Rapidly Increase Near the End of Training.

Why Do We Need Warm-up? A Theoretical Perspective Why Gradients Rapidly Increase Near the End of Training

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.777630Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e1bde069-6bd4-47e7-9f02-85ea680d3b84 · outbound

This paper cites Optimal Linear Decay Learning Rate Schedules and Further Refinements.

Why Do We Need Warm-up? A Theoretical Perspective Optimal Linear Decay Learning Rate Schedules and Further Refinements

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:53.947945Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 184c32fa-1ef5-4c17-a20f-3776cb854a4e · outbound

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

Why Do We Need Warm-up? A Theoretical Perspective An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

Reference 9

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.083987Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a5ed8a55-fe3a-47d2-8dbc-4f922557ea2f · outbound

This paper cites Training Dynamics of the Cooldown Stage in Warmup-Stable-Decay Learning Rate Scheduler.

Why Do We Need Warm-up? A Theoretical Perspective Training Dynamics of the Cooldown Stage in Warmup-Stable-Decay Learning Rate Scheduler

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.214964Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 70b39ca0-925c-4eb8-8faa-2bdc87902b32 · outbound

This paper cites Algorithmic regularization in learning deep homogeneous models: Layers are automatically balanced.

Why Do We Need Warm-up? A Theoretical Perspective Algorithmic regularization in learning deep homogeneous models: Layers are automatically balanced

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.348591Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d1f73df-10a9-4d79-b227-7833a31e70a4 · outbound

This paper cites Beyond uniform smoothness: A stopped analysis of adaptive sgd.

Why Do We Need Warm-up? A Theoretical Perspective Beyond uniform smoothness: A stopped analysis of adaptive sgd

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.532210Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9e70ac17-cdb1-4ac4-b931-0c8ea7a2d36d · outbound

This paper cites Accelerated stochastic optimization methods under quasar-convexity.

Why Do We Need Warm-up? A Theoretical Perspective Accelerated stochastic optimization methods under quasar-convexity

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.667724Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 491f06aa-614d-407b-b324-f5fd86136d8c · outbound

This paper cites Convergence of Clipped SGD on Convex $(L_0,L_1)$-Smooth Functions.

Why Do We Need Warm-up? A Theoretical Perspective Convergence of Clipped SGD on Convex $(L_0,L_1)$-Smooth Functions

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:54.874128Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f3373c31-9333-445c-982e-05503cf50e41 · outbound

This paper cites A Loss Curvature Perspective on Training Instability in Deep Learning.

Why Do We Need Warm-up? A Theoretical Perspective A Loss Curvature Perspective on Training Instability in Deep Learning

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.058339Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 72115b05-738f-407a-ae44-a5cd7453e73a · outbound

This paper cites Methods for Convex $(L_0,L_1)$-Smooth Optimization: Clipping, Acceleration, and Adaptivity.

Why Do We Need Warm-up? A Theoretical Perspective Methods for Convex $(L_0,L_1)$-Smooth Optimization: Clipping, Acceleration, and Adaptivity

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.256221Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation eee57c98-a2a0-4b37-b089-570b82feb3a2 · outbound

This paper cites A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation.

Why Do We Need Warm-up? A Theoretical Perspective A Closer Look at Deep Learning Heuristics: Learning rate restarts, Warmup and Distillation

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.371093Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a24f7417-8014-4bcd-af6c-ac5c3a40258a · outbound

This paper cites Sgd for structured nonconvex functions: Learning rates, minibatching and interpolation.

Why Do We Need Warm-up? A Theoretical Perspective Sgd for structured nonconvex functions: Learning rates, minibatching and interpolation

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.510663Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cdc5061f-6659-4cd0-8063-56f34f4ec4bc · outbound

This paper cites Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour.

Why Do We Need Warm-up? A Theoretical Perspective Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.605650Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f633f650-1203-4683-aafd-08491a57fc17 · outbound

This paper cites No Wrong Turns: The Simple Geometry Of Neural Networks Optimization Paths.

Why Do We Need Warm-up? A Theoretical Perspective No Wrong Turns: The Simple Geometry Of Neural Networks Optimization Paths

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.710545Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 590dcdf8-1c8f-4e78-a4f7-8312681dabf5 · outbound

This paper cites Scaling laws and compute-optimal training beyond fixed training durations.

Why Do We Need Warm-up? A Theoretical Perspective Scaling laws and compute-optimal training beyond fixed training durations

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:55.886801Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation eec30366-a3cc-4b14-a855-8cd81cc3c4a4 · outbound

This paper cites Gradient descent learns linear dynamical systems.

Why Do We Need Warm-up? A Theoretical Perspective Gradient descent learns linear dynamical systems

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.010775Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9da8f203-1a1f-49b6-9256-473c352cb250 · outbound

This paper cites Deep residual learning for image recognition.

Why Do We Need Warm-up? A Theoretical Perspective Deep residual learning for image recognition

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.108598Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b75d76bc-65a8-42f5-b00a-7d9f6a864020 · outbound

This paper cites Gaussian Error Linear Units (GELUs).

Why Do We Need Warm-up? A Theoretical Perspective Gaussian Error Linear Units (GELUs)

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.263519Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 383d99fb-5cdd-4ed3-a254-cb89e9b486d4 · outbound

This paper cites Near-optimal methods for minimizing star-convex functions and beyond.

Why Do We Need Warm-up? A Theoretical Perspective Near-optimal methods for minimizing star-convex functions and beyond

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.359484Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d9a7c06d-9155-4ed8-87d9-c3264157ab62 · outbound

This paper cites Training Compute-Optimal Large Language Models.

Why Do We Need Warm-up? A Theoretical Perspective Training Compute-Optimal Large Language Models

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.463192Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6269db09-27c3-495f-9c95-86dc0e42ccac · outbound

This paper cites An empirical analysis of compute-optimal large language model training.

Why Do We Need Warm-up? A Theoretical Perspective An empirical analysis of compute-optimal large language model training

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.629540Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4fb71426-0151-48e3-ba08-f7b28326a805 · outbound

This paper cites MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies.

Why Do We Need Warm-up? A Theoretical Perspective MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.790434Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 036ce580-74e4-433c-a011-634ce01a186c · outbound

This paper cites Improving transformer optimization through better initialization.

Why Do We Need Warm-up? A Theoretical Perspective Improving transformer optimization through better initialization

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:56.930852Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 00671aad-5bd1-43cb-95ec-42f36a89434a · outbound

This paper cites Loss landscape characterization of neural networks without over-parametrization.

Why Do We Need Warm-up? A Theoretical Perspective Loss landscape characterization of neural networks without over-parametrization

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.081181Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3d011ce7-8726-40bc-91bb-7a331b5e3f86 · outbound

This paper cites Why warmup the learning rate? underlying mechanisms and improvements.

Why Do We Need Warm-up? A Theoretical Perspective Why warmup the learning rate? underlying mechanisms and improvements

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.244841Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation aef0ebda-5f64-4efa-aed9-2b6e38e4f6ad · outbound

This paper cites Linear convergence of gradient and proximal-gradient methods under the polyak- ojasiewicz condition.

Why Do We Need Warm-up? A Theoretical Perspective Linear convergence of gradient and proximal-gradient methods under the polyak- ojasiewicz condition

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.374428Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2b36a278-8f3b-4563-8896-c93683abe061 · outbound

This paper cites an unresolved cited work.

Why Do We Need Warm-up? A Theoretical Perspective Unresolved cited work

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.507142Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 99514f4e-22b3-4f30-957e-afc94d8b474f · outbound

This paper cites Deep learning without poor local minima.

Why Do We Need Warm-up? A Theoretical Perspective Deep learning without poor local minima

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.609783Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 116cd242-7c43-4402-9b22-ea8c50783d42 · outbound

This paper cites Adam: A Method for Stochastic Optimization.

Why Do We Need Warm-up? A Theoretical Perspective Adam: A Method for Stochastic Optimization

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.721743Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 706da65f-d409-42cf-bd5b-c7c9ea81f64f · outbound

This paper cites An alternative view: When does sgd escape local minima? In International conference on machine learning, 2018.

Why Do We Need Warm-up? A Theoretical Perspective An alternative view: When does sgd escape local minima? In International conference on machine learning, 2018

Reference 36

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.819361Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b1228a4f-78df-4eb7-9f23-4784716f46cf · outbound

This paper cites Accelerating SGDM via Learning Rate and Batch Size Schedules: A Lyapunov-Based Analysis.

Why Do We Need Warm-up? A Theoretical Perspective Accelerating SGDM via Learning Rate and Batch Size Schedules: A Lyapunov-Based Analysis

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:57.947144Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a0c02409-bb40-46c1-b44c-3037fcfe5985 · outbound

This paper cites Analyzing & reducing the need for learning rate warmup in gpt training.

Why Do We Need Warm-up? A Theoretical Perspective Analyzing & reducing the need for learning rate warmup in gpt training

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.133345Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b53decd7-bd22-4517-b052-90244089927c · outbound

This paper cites Convex and non-convex optimization under generalized smoothness.

Why Do We Need Warm-up? A Theoretical Perspective Convex and non-convex optimization under generalized smoothness

Reference 39

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.310721Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 719634b7-1aa9-4edb-895a-08254fc9d44c · outbound

This paper cites Loss landscapes and optimization in over-parameterized non-linear systems and neural networks.

Why Do We Need Warm-up? A Theoretical Perspective Loss landscapes and optimization in over-parameterized non-linear systems and neural networks

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.449756Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 21efe338-a1bc-4817-b4e5-3e35f32d45ad · outbound

This paper cites Aiming towards the minimizers: fast convergence of sgd for overparametrized problems.

Why Do We Need Warm-up? A Theoretical Perspective Aiming towards the minimizers: fast convergence of sgd for overparametrized problems

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.650623Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4a969c6c-cba2-4130-8c93-7308ff18d135 · outbound

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

Why Do We Need Warm-up? A Theoretical Perspective On the Variance of the Adaptive Learning Rate and Beyond

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.819676Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 89caaa2e-7bee-452a-9079-db1bb129da9e · outbound

This paper cites Theoretical Analysis on how Learning Rate Warmup Accelerates Convergence.

Why Do We Need Warm-up? A Theoretical Perspective Theoretical Analysis on how Learning Rate Warmup Accelerates Convergence

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:58.891916Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 648c8223-3bce-474f-ad61-e981523edfb9 · outbound

This paper cites SGDR: Stochastic Gradient Descent with Warm Restarts.

Why Do We Need Warm-up? A Theoretical Perspective SGDR: Stochastic Gradient Descent with Warm Restarts

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.029439Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 88d91fec-cb0e-47cd-adb3-84a865af0838 · outbound

This paper cites Decoupled Weight Decay Regularization.

Why Do We Need Warm-up? A Theoretical Perspective Decoupled Weight Decay Regularization

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.193626Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a1bac448-bfbd-4ba4-b4da-9731b4c89095 · outbound

This paper cites The power of interpolation: Understanding the effectiveness of sgd in modern over-parametrized learning.

Why Do We Need Warm-up? A Theoretical Perspective The power of interpolation: Understanding the effectiveness of sgd in modern over-parametrized learning

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.249809Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ffa64622-a0a7-40a3-a49a-eeea6c54b6bd · outbound

This paper cites Matrix differential calculus with applications to simple, hadamard, and kronecker products.

Why Do We Need Warm-up? A Theoretical Perspective Matrix differential calculus with applications to simple, hadamard, and kronecker products

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.350720Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ed497575-5314-43e2-91da-05d9b6c4f757 · outbound

This paper cites An Empirical Model of Large-Batch Training.

Why Do We Need Warm-up? A Theoretical Perspective An Empirical Model of Large-Batch Training

Reference 48

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.426040Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dcbb8440-442b-4c34-9222-c176b5e1dabb · outbound

This paper cites The fineweb datasets: Decanting the web for the finest text data at scale.

Why Do We Need Warm-up? A Theoretical Perspective The fineweb datasets: Decanting the web for the finest text data at scale

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.495969Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 75c04dd8-88c7-448a-bd6c-7b29f0a42f71 · outbound

This paper cites Gradient methods for minimizing functionals.

Why Do We Need Warm-up? A Theoretical Perspective Gradient methods for minimizing functionals

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.578629Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5b7c22af-6cb4-40cd-b65c-e90f2ac141be · outbound

This paper cites Language models are unsupervised multitask learners.

Why Do We Need Warm-up? A Theoretical Perspective Language models are unsupervised multitask learners

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.674564Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 95a62f5f-478d-46fb-89b6-07d1f2248832 · outbound

This paper cites Gluon: Making Muon & Scion Great Again! (Bridging Theory and Practice of LMO-based Optimizers for LLMs).

Why Do We Need Warm-up? A Theoretical Perspective Gluon: Making Muon & Scion Great Again! (Bridging Theory and Practice of LMO-based Optimizers for LLMs)

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.776219Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 05b0d5fe-f2e7-4478-b1b0-2521860b2175 · outbound

This paper cites Stepping on the edge: Curvature aware learning rate tuners.

Why Do We Need Warm-up? A Theoretical Perspective Stepping on the edge: Curvature aware learning rate tuners

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.901620Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ce80e345-2c2a-46c4-b820-4fbd67f88085 · outbound

This paper cites The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training.

Why Do We Need Warm-up? A Theoretical Perspective The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-04T12:38:59.978009Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e558cf7f-ddfc-4c8a-a555-a88298d98e2f · outbound

This paper cites GLU Variants Improve Transformer.

Why Do We Need Warm-up? A Theoretical Perspective GLU Variants Improve Transformer

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.127984Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d5a15b6-fd2b-4aba-8e62-141cba87251f · outbound

This paper cites On the generalization benefit of noise in stochastic gradient descent.

Why Do We Need Warm-up? A Theoretical Perspective On the generalization benefit of noise in stochastic gradient descent

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.284259Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2ba07a3e-b380-47db-96df-167002843834 · outbound

This paper cites Roformer: Enhanced transformer with rotary position embedding.

Why Do We Need Warm-up? A Theoretical Perspective Roformer: Enhanced transformer with rotary position embedding

Reference 57

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.425116Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 217d4747-d2a8-4904-a697-80d9c57895c1 · outbound

This paper cites On the importance of initialization and momentum in deep learning.

Why Do We Need Warm-up? A Theoretical Perspective On the importance of initialization and momentum in deep learning

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.578854Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4b7c4293-5cfe-4b91-af61-bbeaacf0670a · outbound

This paper cites Fast convergence in learning two-layer neural networks with separable data.

Why Do We Need Warm-up? A Theoretical Perspective Fast convergence in learning two-layer neural networks with separable data

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.714356Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation adeb38bf-a005-40ec-8257-9ae372c03ed9 · outbound

This paper cites Llama 2: Open Foundation and Fine-Tuned Chat Models.

Why Do We Need Warm-up? A Theoretical Perspective Llama 2: Open Foundation and Fine-Tuned Chat Models

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:00.861463Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 84f68236-ec75-4826-868b-f8c0866214bc · outbound

This paper cites Empirical tests of optimization assumptions in deep learning.

Why Do We Need Warm-up? A Theoretical Perspective Empirical tests of optimization assumptions in deep learning

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.030025Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a80581cf-2114-4a12-88fb-cf5e581d546b · outbound

This paper cites Optimizing $(L_0, L_1)$-Smooth Functions by Gradient Methods.

Why Do We Need Warm-up? A Theoretical Perspective Optimizing $(L_0, L_1)$-Smooth Functions by Gradient Methods

Reference 62

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.193581Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a1650146-ef7d-4966-b190-4e8aa5da6398 · outbound

This paper cites Attention is all you need.

Why Do We Need Warm-up? A Theoretical Perspective Attention is all you need

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.314115Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5f300618-eb15-4eee-8bcf-2c4bf1534160 · outbound

This paper cites Convergence of adagrad for non-convex objectives: Simple proofs and relaxed assumptions.

Why Do We Need Warm-up? A Theoretical Perspective Convergence of adagrad for non-convex objectives: Simple proofs and relaxed assumptions

Reference 65

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.594610Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc431704-a053-41e8-993a-0db5dd316058 · outbound

This paper cites Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective.

Why Do We Need Warm-up? A Theoretical Perspective Understanding Warmup-Stable-Decay Learning Rates: A River Valley Loss Landscape Perspective

Reference 66

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.733820Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 125f1764-fac9-4207-85eb-4803f87867a0 · outbound

This paper cites Small-scale proxies for large-scale Transformer training instabilities.

Why Do We Need Warm-up? A Theoretical Perspective Small-scale proxies for large-scale Transformer training instabilities

Reference 67

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:01.892794Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a98b8919-b85a-4f15-abff-ca0412f2a3df · outbound

This paper cites On the overlooked pitfalls of weight decay and how to mitigate them: A gradient-norm perspective.

Why Do We Need Warm-up? A Theoretical Perspective On the overlooked pitfalls of weight decay and how to mitigate them: A gradient-norm perspective

Reference 68

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.101607Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9abf9a17-3192-4b9a-abf4-93183a4fca72 · outbound

This paper cites On layer normalization in the transformer architecture.

Why Do We Need Warm-up? A Theoretical Perspective On layer normalization in the transformer architecture

Reference 69

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.305117Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1e455ef0-d364-4d76-a416-edcf5f517509 · outbound

This paper cites Dive into deep learning.

Why Do We Need Warm-up? A Theoretical Perspective Dive into deep learning

Reference 70

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.459232Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a2fe7e7d-00db-4cad-840f-f698d4c32a39 · outbound

This paper cites Root mean square layer normalization.

Why Do We Need Warm-up? A Theoretical Perspective Root mean square layer normalization

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.605680Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation dcdfc5bf-6ca6-49af-94f2-6e3622e0ceda · outbound

This paper cites Improved analysis of clipping algorithms for non-convex optimization.

Why Do We Need Warm-up? A Theoretical Perspective Improved analysis of clipping algorithms for non-convex optimization

Reference 72

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.697714Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b8c02199-f48d-4b06-a70f-52f3feac5a49 · outbound

This paper cites Why gradient clipping accelerates training: A theoretical justification for adaptivity.

Why Do We Need Warm-up? A Theoretical Perspective Why gradient clipping accelerates training: A theoretical justification for adaptivity

Reference 73

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.842307Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d90af45d-98dc-4e8a-9d6f-b03a9e6d6e45 · outbound

This paper cites On the convergence and improvement of stochastic normalized gradient descent.

Why Do We Need Warm-up? A Theoretical Perspective On the convergence and improvement of stochastic normalized gradient descent

Reference 74

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:02.990431Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ba8c39b4-7f2b-47c8-aa2b-6a096121e6db · outbound

This paper cites @esa (Ref.

Why Do We Need Warm-up? A Theoretical Perspective @esa (Ref

Reference 75

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:03.122139Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 28aafa4e-8ded-42f3-a3f4-0ad38030173d · outbound

This paper cites an unresolved cited work.

Why Do We Need Warm-up? A Theoretical Perspective Unresolved cited work

Reference 76

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:03.316608Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c689fca2-05c8-4d5c-bb73-37bdf0933cbd · outbound

This paper cites an unresolved cited work.

Why Do We Need Warm-up? A Theoretical Perspective Unresolved cited work

Reference 77

Resolution: unresolved; no resolver link, observed 2026-08-04T12:39:03.472216Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 91e5ed69-2867-42ed-bbba-e29abcf80be0 · inbound

Self-Play Enhancement via Advantage-Weighted Refinement in Online Federated LLM Fine-Tuning with Real-Time Feedback cites this paper.

Self-Play Enhancement via Advantage-Weighted Refinement in Online Federated LLM Fine-Tuning with Real-Time Feedback Why Do We Need Warm-up? A Theoretical Perspective

Reference 1

Resolution: verified exact; arxiv_id, observed 2026-06-30T02:16:10.224205Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.

Observation c2ce85ae-8831-4397-8cca-f6e072d2affe · inbound

Avoiding Bias in Clipped SGD for Overparameterized Models under Generalized Smoothness cites this paper.

Avoiding Bias in Clipped SGD for Overparameterized Models under Generalized Smoothness Why Do We Need Warm-up? A Theoretical Perspective

Reference 2

Resolution: verified exact; local_arxiv, observed 2026-06-30T20:15:04.538172Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.