Paper Citation Record · LEDGER

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

As of 10 August 2026, this Paper Citation Record lists 30 of 30 outbound references and 8 inbound Pith citation observations for arXiv:2501.18965.

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

pith.paper-citation-record.v1
2501.18965 v2

Coverage vector

measured 30 of 30 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-09T21:56:50.876409Z

measured 38 of 38 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-10T06:31:04.303077+00:00

measured 8 of 8 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-06T20:48:53.764373Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-05-25T05:40:24.434095Z

Reference resolution

30 of 30 outbound references displayed

verified exact1
verified fuzzy12
unresolved15
parse uncertain0
malformed identifier2
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 6dd21981-5183-4264-a3d5-78c4a7bbfeeb · outbound

This paper cites Why you don't overfit, and don't need Bayes if you only train for one epoch.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Why you don't overfit, and don't need Bayes if you only train for one epoch

Reference 1

Resolution: verified exact; local_arxiv, observed 2026-08-09T21:56:51.068688Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 679b2f08-7bae-4f38-a898-433e752cb21b · outbound

This paper cites For validation set metrics, we display a running average over five epoch in thick to smoothen the plot, and the original data in thin.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training For validation set metrics, we display a running average over five epoch in thick to smoothen the plot, and the original data in thin

Reference 2

Resolution: malformed identifier; raw_fallback, observed 2026-08-09T21:56:51.164759Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation a3b291a1-492c-4fff-8e31-19e824404fff · outbound

This paper cites Loss Landscape Characterization of Neural Networks without Over-Parametrization.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Loss Landscape Characterization of Neural Networks without Over-Parametrization

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.773093Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 63c08d1b-7f65-464c-9cc5-dde6561344ff · outbound

This paper cites Rethinking Conventional Wisdom in Machine Learning: From Generalization to Scaling.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Rethinking Conventional Wisdom in Machine Learning: From Generalization to Scaling

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.797601Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 16e15bbc-102e-42d2-ac99-0f888910f114 · outbound

This paper cites No More Adam: Learning Rate Scaling at Initialization is All You Need.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training No More Adam: Learning Rate Scaling at Initialization is All You Need

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.802232Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3e8e6c92-2518-48cc-b0f9-d75f5f17155d · outbound

This paper cites Exact convergence rate of the last iterate in subgradient methods.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Exact convergence rate of the last iterate in subgradient methods

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.806915Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8b6da73c-238a-49bb-b9df-bb02ea17d8e7 · outbound

This paper cites • Appendix B: supplementary information on our experiments.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training • Appendix B: supplementary information on our experiments

Reference 15

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.305651Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 89f7914a-b082-4f20-847d-b31863ba5220 · outbound

This paper cites 2, but with Ωt from (11) The bound on the best-so-far bound has a very different shape of the last-iterate bound.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training 2, but with Ωt from (11) The bound on the best-so-far bound has a very different shape of the last-iterate bound

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.290719Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation e643bf3e-f10b-4357-b32d-ae064670ba1c · outbound

This paper cites Convergence is plotted with the optimal base learning-rate γ⋆ (chosen individually for each schedule).

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Convergence is plotted with the optimal base learning-rate γ⋆ (chosen individually for each schedule)

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.261522Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 0ed11f91-b193-4379-9e99-d2f2333c5d1a · outbound

This paper cites Compare to Figure A1 in Hoffmann et al.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Compare to Figure A1 in Hoffmann et al

Reference 19

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.246198Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1f0b1871-10d9-4c91-a674-e23fe81577af · outbound

This paper cites Details on Experiments in Fig.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Details on Experiments in Fig

Reference 20

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.232415Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation aa67bb27-c4a6-4553-a890-8a8ebdc4f619 · outbound

This paper cites Dark grey marks the bound of the constant schedule.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Dark grey marks the bound of the constant schedule

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.192616Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 70898871-83c4-4d1c-a77e-69c6cef1d03e · outbound

This paper cites We observe the same characteristic drop of the loss for wsd, as well as matching performance of wsd and cosine.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training We observe the same characteristic drop of the loss for wsd, as well as matching performance of wsd and cosine

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.148945Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation b1d64691-5d65-47b3-a0a8-76e3c4edb4e6 · outbound

This paper cites an unresolved cited work.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Unresolved cited work

Reference 28

Resolution: unresolved; raw_fallback, observed 2026-08-09T21:56:51.132093Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 96b1f77e-41a0-4eca-b531-c2ed1d4c5045 · outbound

This paper cites an unresolved cited work.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Unresolved cited work

Reference 29

Resolution: unresolved; raw_fallback, observed 2026-08-09T21:56:51.116057Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5ecfb48b-b270-45d3-b404-f23bf64a44fb · outbound

This paper cites Theorem 3.1 follows from applying Theorem E.2 with ˆηt := γηt.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Theorem 3.1 follows from applying Theorem E.2 with ˆηt := γηt

Reference 30

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.099589Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 5dcbfc1e-52df-47dc-a6ef-1c0aa8709ad4 · outbound

This paper cites From the (generalized) Cauchy-Schwarz inequality combined with Young’s inequality, we have s3 ≤ µ 2 ∥xt+1 − xt∥2 + η2 t 2µ ∥gt∥2 ∗.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training From the (generalized) Cauchy-Schwarz inequality combined with Young’s inequality, we have s3 ≤ µ 2 ∥xt+1 − xt∥2 + η2 t 2µ ∥gt∥2 ∗

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.084392Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation bc4820f7-d152-4063-b115-95492681ba8d · outbound

This paper cites an unresolved cited work.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Unresolved cited work

Reference 400

Resolution: unresolved; raw_fallback, observed 2026-08-09T21:56:51.276044Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 40e4be6d-56a9-4089-8150-8e11153d32d5 · outbound

This paper cites Training trajectories, mini-batch losses and the curious role of the learning rate.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Training trajectories, mini-batch losses and the curious role of the learning rate

Reference 1970

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.782654Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3915129f-23cd-42b6-9143-eec1894a883f · outbound

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

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Optimal Linear Decay Learning Rate Schedules and Further Refinements

Reference 1983

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.758379Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8d100234-d5ee-4cf5-ae7b-91a8a8b4af6a · outbound

This paper cites Chinchilla Scaling: A replication attempt.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Chinchilla Scaling: A replication attempt

Reference 2003

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.749126Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8840c951-3c13-4241-86cc-eb1f689b32d8 · outbound

This paper cites Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler

Reference 2013

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.787977Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5604a1e5-f2ac-4fce-8373-30cd4b0d2f00 · outbound

This paper cites We train all models with SGD with heavy-ball momentum.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training We train all models with SGD with heavy-ball momentum

Reference 2016

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.179031Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 163e6581-3c76-4b8c-9004-a3c6b2493bb0 · outbound

This paper cites Deep residual learning for image recognition.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Deep residual learning for image recognition

Reference 2018

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.768440Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 79b3def1-f6fd-4824-bf96-ced9b94521f0 · outbound

This paper cites For all further details we refer to Hägele et al.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training For all further details we refer to Hägele et al

Reference 2019

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.218222Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 4dbf9e95-152d-4ef3-8b95-50e5dd62eaca · outbound

This paper cites Rockafellar, R.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Rockafellar, R

Reference 2020

Resolution: verified fuzzy; raw_fallback, observed 2026-08-09T21:56:51.319527Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 27202c2b-dd3a-46fe-81e5-8d96359ea414 · outbound

This paper cites an unresolved cited work.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Unresolved cited work

Reference 2021

Resolution: unresolved; raw_fallback, observed 2026-08-09T21:56:51.344700Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation c47b089a-3349-4086-9f92-679beef3cf3d · outbound

This paper cites More concretely, assume we have trained a model of sizeN1 for D1 tokens.

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training More concretely, assume we have trained a model of sizeN1 for D1 tokens

Reference 2022

Resolution: malformed identifier; raw_fallback, observed 2026-08-09T21:56:51.205598Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation b818474e-760d-41e5-b3cd-ab77679033ef · outbound

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

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training LLaMA: Open and Efficient Foundation Language Models

Reference 2023

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.792621Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2e7eb3c5-c3a8-4d26-a2c4-dac886695c21 · outbound

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

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-09T21:56:50.763500Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 524ba4f5-5eea-4805-bd48-9394a5ddabe6 · inbound

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks cites this paper.

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-06T20:48:53.764373Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

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

Why Do We Need Warm-up? A Theoretical Perspective cites this paper.

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 cf95fe02-b66a-4973-9e6e-e17359c23ff9 · inbound

Optimistic Dual Averaging Unifies Modern Optimizers cites this paper.

Optimistic Dual Averaging Unifies Modern Optimizers The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-05-13T05:52:21.853745Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 1b620e58-78fa-470b-81b9-c9e4586f43a6 · inbound

Anytime Training with Schedule-Free Spectral Optimization cites this paper.

Anytime Training with Schedule-Free Spectral Optimization The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 56

Resolution: verified exact; arxiv_id, observed 2026-05-25T05:40:24.438563Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.

Observation 2b84bfe1-2116-411a-b4fe-13fde4b8f7f5 · inbound

WSqD: A Horizon-Free Learning Rate Schedule for Large Model Training cites this paper.

WSqD: A Horizon-Free Learning Rate Schedule for Large Model Training The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 53

Resolution: unresolved; no resolver link, observed 2026-07-14T08:04:06.432613Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4c08e596-2b3d-4b4c-bc89-faa5e206c143 · inbound

AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model cites this paper.

AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 80

Resolution: unresolved; no resolver link, observed 2026-08-01T12:16:36.364978Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c397482a-d1dc-4a09-a687-86898400f35b · inbound

AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model cites this paper.

AGNFormer I: Reconstruction of AGN spectra using a probabilistic transformer model The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 154

Resolution: unresolved; no resolver link, observed 2026-08-01T12:16:45.602773Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 61cc0a2b-020d-4971-aca3-9a3efae113db · inbound

A Defense of the Quadratic Model cites this paper.

A Defense of the Quadratic Model The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-01T06:58:11.864874Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.