Pith. sign in

Paper Citation Record · LEDGER

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes

As of 20 August 2026, this Paper Citation Record lists 83 of 83 outbound references and 2 inbound Pith citation observations for arXiv:2508.00180.

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

pith.paper-citation-record.v1
2508.00180 v1

Coverage vector

measured 83 of 83 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-06T10:23:52.037279Z

measured 85 of 85 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 2 of 2 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-07-01T06:36:46.924135Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-07-04T16:49:58.438161Z

Reference resolution

83 of 83 outbound references displayed

  • verified exact0
  • verified fuzzy33
  • unresolved50
  • parse uncertain0
  • malformed identifier0
  • metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 77d65f5c-2ad6-45b1-8128-7fa0df78b51b · outbound

This paper cites Program Synthesis with Large Language Models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Program Synthesis with Large Language Models

Reference 1

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:45.651809Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:45.651809Z digest=sha256:16a62bbb601874d82e22b0a4af077b6d4e69b2d9303338f0e77f44afd36a82d3

Observation 2fd7dd10-ca30-4ebc-995a-d843395cfb1d · outbound

This paper cites High-dimensional limit theorems for sgd: Effective dynamics and critical scaling.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes High-dimensional limit theorems for sgd: Effective dynamics and critical scaling

Reference 2

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:45.744826Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:45.744826Z digest=sha256:ae86cceca80bb8f639892b9f877b9c60a3e3a7606e44a409ae876d96aedfcd73

Observation 896df6f7-d37c-412f-96ea-fef7358e4f68 · outbound

This paper cites Provable guarantees for generative behavior cloning: Bridging low-level stability and high-level behavior.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Provable guarantees for generative behavior cloning: Bridging low-level stability and high-level behavior

Reference 3

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:45.821133Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:45.821133Z digest=sha256:4cd49d52e102298a4ac582f78282849fb365747fd90f7bd2e24b58252d97c34d

Observation 93c7e947-9878-47dd-9ea3-2962ab80e862 · outbound

This paper cites Butterfly effects of sgd noise: Error amplification in behavior cloning and autoregression.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Butterfly effects of sgd noise: Error amplification in behavior cloning and autoregression

Reference 4

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:45.847970Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:45.847970Z digest=sha256:a0e839add192cc0d34ac401ed28ad0b0acb8139448d66fe8ea5bb91021b24042

Observation 4f3e6366-e1a1-40cf-96b8-f73a2e26bc03 · outbound

This paper cites Approximation methods which converge with probability one.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Approximation methods which converge with probability one

Reference 5

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.713434Z

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.

source=arxiv_source observed=2026-08-06T10:23:45.891086Z digest=sha256:0877ecee18c304073edfa75bb7554017c0df8173590b3a9b0e58845358c61fc9

Observation 71712a5f-ec88-4c23-b972-f6cb2bfc0284 · outbound

This paper cites How to scale your ema.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes How to scale your ema

Reference 6

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.704441Z

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.

source=arxiv_source observed=2026-08-06T10:23:46.086657Z digest=sha256:a57eb6a3eacd94b993936eb357d0286944ef75563ed3bb83d9f637c629948f3e

Observation 739fa8a4-4a7e-4faf-9323-205d53801095 · outbound

This paper cites LEGAL-BERT: The Muppets straight out of Law School.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes LEGAL-BERT: The Muppets straight out of Law School

Reference 7

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.241953Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.241953Z digest=sha256:834f08c4dc3944820132df35644d6e936b1f15bd90d479ab971ccd5dc859d3ee

Observation 658e397c-b1a2-46e6-a480-ba086aed9129 · outbound

This paper cites Incorrect baseline evaluations call into question recent llm-rl claims, 2025.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Incorrect baseline evaluations call into question recent llm-rl claims, 2025

Reference 8

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.694876Z

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.

source=arxiv_source observed=2026-08-06T10:23:46.351256Z digest=sha256:c2d6bbed8993fffeebe39c6b9083fc5b742f4c5c22fba9b0a19832e58ed566c5

Observation 1693cfb4-1984-47db-a98a-c47900d5666c · outbound

This paper cites Learning to Generate Better Than Your LLM.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Learning to Generate Better Than Your LLM

Reference 9

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.465760Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.465760Z digest=sha256:9be99eceb6fd196f09a6abb7e645e89c94bfa1dcfa3d3bb733fa294f81e82b80

Observation e92d7ade-133c-4a6c-a0a6-5c32c8cc73db · outbound

This paper cites Bidirectional looking with a novel double exponential moving average to adaptive and non-adaptive momentum optimizers.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Bidirectional looking with a novel double exponential moving average to adaptive and non-adaptive momentum optimizers

Reference 10

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.685663Z

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.

source=arxiv_source observed=2026-08-06T10:23:46.525397Z digest=sha256:36b933b95d875bd454d2cad91ca9854fd4ad646a32a93ba3b8d2382073e35700

Observation 62db7694-fe61-409d-9657-a56932319d04 · outbound

This paper cites Double/debiased machine learning for treatment and structural parameters, 2018.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Double/debiased machine learning for treatment and structural parameters, 2018

Reference 11

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.595345Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.595345Z digest=sha256:1d780de0f9afb3ed37aa724e41a8d097263ed8f65b1097d237e198f4323b70ac

Observation 8083e651-f737-47e3-8a4c-7f1cb04ceaeb · outbound

This paper cites Deep reinforcement learning from human preferences.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Deep reinforcement learning from human preferences

Reference 12

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.712637Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.712637Z digest=sha256:b43d799f8b84521387b39b47a0197736c8510e66bd949dd8270c509b77750b7e

Observation e6e0a726-cff1-4a2a-b042-9c8d44377b3c · outbound

This paper cites BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions

Reference 13

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.829935Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.829935Z digest=sha256:951ffdda1721272674cc3ab75002b14fc2d860e647fe13c6a1465103486cdc1f

Observation cfa83e9a-5329-42f5-a3fc-6079bdc55943 · outbound

This paper cites Training Verifiers to Solve Math Word Problems.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Training Verifiers to Solve Math Word Problems

Reference 14

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.868383Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.868383Z digest=sha256:069ed43731ec51231594b5ff55c765e089e3512745c052f8c12531a6432b016c

Observation 9d96bd46-3797-412d-a50d-1b9585d41287 · outbound

This paper cites Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Gradient Descent on Neural Networks Typically Occurs at the Edge of Stability

Reference 15

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:46.940838Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:46.940838Z digest=sha256:67480833e51e00bf16564aa0c15ad5c7dbafd508e36a31c96c5b3cf3eef5c76a

Observation 39618f0a-9a58-4c81-92e1-093b7557d8d4 · outbound

This paper cites Saga: A fast incremental gradient method with support for non-strongly convex composite objectives.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Saga: A fast incremental gradient method with support for non-strongly convex composite objectives

Reference 16

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.049340Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.049340Z digest=sha256:c94f5536879ef7fc2f37b026794c55d3ca37a51c0b47c7da776caa40d662ceac

Observation 289142f1-3423-4b30-9061-e1b4c64b5796 · outbound

This paper cites Averaged least-mean-squares: Bias-variance trade-offs and optimal sampling distributions.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Averaged least-mean-squares: Bias-variance trade-offs and optimal sampling distributions

Reference 17

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.660256Z

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.

source=arxiv_source observed=2026-08-06T10:23:47.109317Z digest=sha256:2e4c2e243278347852d0e64bb7f5b05ee83b1970ffffa407534781d2a6c8fe9a

Observation ccb16d40-0472-4d07-b715-80d9f10420f0 · outbound

This paper cites Harder, better, faster, stronger convergence rates for least-squares regression.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Harder, better, faster, stronger convergence rates for least-squares regression

Reference 18

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.650886Z

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.

source=arxiv_source observed=2026-08-06T10:23:47.232493Z digest=sha256:d7026b61fb87a98b5f98991c9479d619f955ae77d797d56017062c9f56e9e8ef

Observation 13b7bb9e-5939-40de-b10a-ca467174d357 · outbound

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

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Adaptive subgradient methods for online learning and stochastic optimization

Reference 19

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.286455Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.286455Z digest=sha256:d64dcb52523630015d1f444a50fd1611214ae30b360d9a36aa7b74e0f08f9172

Observation 5dfb3620-86a9-4642-8b03-f979ba971b2f · outbound

This paper cites Is behavior cloning all you need? understanding horizon in imitation learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Is behavior cloning all you need? understanding horizon in imitation learning

Reference 20

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.636518Z

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.

source=arxiv_source observed=2026-08-06T10:23:47.382502Z digest=sha256:066007f5488a35aa1cf9e26009714d6bcd2c0644e9130cd70e50e4f8d5b7ac7c

Observation 98952c78-f52a-43a7-8762-b8befb25b1ea · outbound

This paper cites The Llama 3 Herd of Models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes The Llama 3 Herd of Models

Reference 21

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.454252Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.454252Z digest=sha256:87d1c2bc9bbecb9ae11b03dc98130197acaf689714ae1171f63370e0246b41b5

Observation 2d917c7c-1e02-4cba-b043-0fc3e628a464 · outbound

This paper cites Bootstrap your own latent-a new approach to self-supervised learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Bootstrap your own latent-a new approach to self-supervised learning

Reference 22

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.628324Z

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.

source=arxiv_source observed=2026-08-06T10:23:47.588289Z digest=sha256:e3dc664307c5d28ed8bb2471dcccf6859dd0e6ad28bd441f6b6459c721f55889

Observation c7d64f01-90ce-482c-9e60-ac4c9f22bc6d · outbound

This paper cites Shampoo: Preconditioned stochastic tensor optimization.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Shampoo: Preconditioned stochastic tensor optimization

Reference 23

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.675863Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.675863Z digest=sha256:635ec17d0e2c736481e8b9ed623c9ad7720e736255148d093dc7523def4d64e4

Observation b5a3b82a-2d64-46fe-9b46-5323414f16b9 · outbound

This paper cites Measuring massive multitask language understanding.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Measuring massive multitask language understanding

Reference 24

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.715586Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.715586Z digest=sha256:831248759553005fab68813fe73202450965b06d62db7da890855ae549185031

Observation 25f37361-a597-4e58-b341-fbe7cadf4fe0 · outbound

This paper cites Averaging Weights Leads to Wider Optima and Better Generalization.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Averaging Weights Leads to Wider Optima and Better Generalization

Reference 25

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.788636Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.788636Z digest=sha256:4e8e87a25bbaf4d6ff5e268e2c4807b5344aa571470779c66bbdcef2822cc5cf

Observation 2b320dce-cf3e-4c70-ac64-2d6cafb1a769 · outbound

This paper cites Neural tangent kernel: Convergence and generalization in neural networks.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Neural tangent kernel: Convergence and generalization in neural networks

Reference 26

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:47.969468Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:47.969468Z digest=sha256:f452775d7e251c766130297f0bd0e05ff39671c3770e1fb03c2d684b996addb1

Observation 1896506d-1216-420a-8999-dadc4a973a61 · outbound

This paper cites Accelerating stochastic gradient descent using predictive variance reduction.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Accelerating stochastic gradient descent using predictive variance reduction

Reference 27

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:48.111032Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:48.111032Z digest=sha256:3b3d0720c14f55a4c75a7fa2a4abf2f8a0ec4cf436e44e677ad49d2cf448026a

Observation 96672192-4172-4668-8c19-b26010b8810b · outbound

This paper cites Stop Wasting My Time! Saving Days of ImageNet and BERT Training with Latest Weight Averaging.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Stop Wasting My Time! Saving Days of ImageNet and BERT Training with Latest Weight Averaging

Reference 28

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:48.286379Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:48.286379Z digest=sha256:d57fb5388c4e753c361b439c405d7df59dbe6fae53a1ddbbd95f00516e69a1e3

Observation 09d638c1-2edc-4312-bfa7-39834e488387 · outbound

This paper cites No train no gain: Revisiting efficient training algorithms for transformer-based language models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes No train no gain: Revisiting efficient training algorithms for transformer-based language models

Reference 29

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.594443Z

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.

source=arxiv_source observed=2026-08-06T10:23:48.404641Z digest=sha256:3cda9401945fc55182e6f82bf99b568a02061d8cab363bfb225d60de84e01d33

Observation 27489973-0d4f-4fca-bbb8-1ddfbc324af7 · outbound

This paper cites Gemma 3 Technical Report.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Gemma 3 Technical Report

Reference 30

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:48.550625Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:48.550625Z digest=sha256:dcdf08673072b2aa57ce8269eddeaeb5017db7ef42e0b77d1f2e60a9ceeff9c4

Observation edf4b9f7-060b-40ef-b8ac-f239fde50a20 · outbound

This paper cites Analyzing and Improving the Training Dynamics of Diffusion Models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Analyzing and Improving the Training Dynamics of Diffusion Models

Reference 31

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:48.659364Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:48.659364Z digest=sha256:578fe0b839d74dc0aa1edada9cd4210d837b4388770be9e7d3ab08c1ef337791

Observation 32dc7719-8ef9-4931-9a68-79d2831431ec · outbound

This paper cites Adam: A Method for Stochastic Optimization.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Adam: A Method for Stochastic Optimization

Reference 32

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:48.814561Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:48.814561Z digest=sha256:9f1ffa8e29f4131eee39b75e49fe2409678b7d2031c8d8799dcc41aa05815932

Observation 8798bd86-4e58-4e5f-b18c-fc4f255e4fee · outbound

This paper cites an unresolved cited work.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Unresolved cited work

Reference 33

Resolution
unresolved
raw_fallback, observed 2026-08-06T10:23:52.585249Z

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.

source=arxiv_source observed=2026-08-06T10:23:48.845871Z digest=sha256:803a346d702f31068359ecfac0a793ef8eb96bc6adb4704810d9f65358d347b9

Observation a4e38958-5226-4474-ad47-caaca1076791 · outbound

This paper cites Statistical inference for ergodic diffusion processes.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Statistical inference for ergodic diffusion processes

Reference 34

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.576321Z

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.

source=arxiv_source observed=2026-08-06T10:23:48.958790Z digest=sha256:024b0bb8f851330023711fee051545d8cf6108d476e0fc01e9904fa938ae7911

Observation c6e724a6-2d50-4fcb-8afe-e57fca99404a · outbound

This paper cites Gonzalez, Hao Zhang, and Ion Stoica.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Gonzalez, Hao Zhang, and Ion Stoica

Reference 35

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:49.190251Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:49.190251Z digest=sha256:d9119673e46ed83b0deb955e77603808265259d875ae5c238a1536ca5f779357

Observation a0a7993c-f5cd-4937-9f93-9d7ed7c6069b · outbound

This paper cites Tulu 3: Pushing Frontiers in Open Language Model Post-Training.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Tulu 3: Pushing Frontiers in Open Language Model Post-Training

Reference 36

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:49.237889Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:49.237889Z digest=sha256:b9d1a4214826f6157943fd1f947f0d327fc89570e37b29d6100e06725ad8562c

Observation 66afda0a-b3f5-4f90-ae01-639a92707c1b · outbound

This paper cites Brownian motion, martingales, and stochastic calculus.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Brownian motion, martingales, and stochastic calculus

Reference 37

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.561802Z

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.

source=arxiv_source observed=2026-08-06T10:23:49.371231Z digest=sha256:af3785cbe48331d1998bbce6a7919971450febea1705595daf5cd0bda472f860

Observation f31e223d-ed04-4978-b17c-ea41dc475b74 · outbound

This paper cites Slow and Steady Wins the Race: Maintaining Plasticity with Hare and Tortoise Networks.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Slow and Steady Wins the Race: Maintaining Plasticity with Hare and Tortoise Networks

Reference 38

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:49.456993Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:49.456993Z digest=sha256:51c52154e2ae2e9b0166fd8abf4f774fb320bc714a04979061c1742fa076c2c6

Observation 9bf4a260-eafb-4224-a215-816eb2051fb1 · outbound

This paper cites Biobert: a pre-trained biomedical language representation model for biomedical text mining.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Biobert: a pre-trained biomedical language representation model for biomedical text mining

Reference 39

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:49.577990Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:49.577990Z digest=sha256:ed534bfc98503909a3d2dd17327b1d53bdb18e8e98951d120056854c7a1211a8

Observation 02ffbe8d-dd24-4b18-a2f7-d2fa7f8dec37 · outbound

This paper cites Theory of point estimation.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Theory of point estimation

Reference 40

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.548229Z

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.

source=arxiv_source observed=2026-08-06T10:23:49.729624Z digest=sha256:759cf680f625acb4107a264dc701e0a84104917d17c96650a45a1f29b75c497a

Observation 00607ca9-db1b-4724-b9f2-5151132513de · outbound

This paper cites Stochastic modified equations and adaptive stochastic gradient algorithms.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Stochastic modified equations and adaptive stochastic gradient algorithms

Reference 41

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.539762Z

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.

source=arxiv_source observed=2026-08-06T10:23:49.847926Z digest=sha256:e82766951a6a7335bd4282fdc2bbb1c5c498e02601b2ad6a4f620343b66f7545

Observation 314f98bd-a0a5-47c9-86d3-210482100d3b · outbound

This paper cites Switch EMA: A Free Lunch for Better Flatness and Sharpness.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Switch EMA: A Free Lunch for Better Flatness and Sharpness

Reference 42

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:49.973613Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:49.973613Z digest=sha256:c7bd298f7e6ace928df2d4ad866a0f896e327e63b24d3df7c83fb1ddcb333c3a

Observation 6715869e-a04f-4124-85bc-61f40ef76e69 · outbound

This paper cites Statistics of random processes: I.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Statistics of random processes: I

Reference 43

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.529338Z

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.

source=arxiv_source observed=2026-08-06T10:23:50.040897Z digest=sha256:823290475c75bf84e5e5de34f99c192a2cc23cae491312a52fe3bf93cf8279b2

Observation eb64e063-3c2a-47b7-b32f-7ce796c46572 · outbound

This paper cites Statistics of random processes II: Applications, volume 6.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Statistics of random processes II: Applications, volume 6

Reference 44

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.520280Z

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.

source=arxiv_source observed=2026-08-06T10:23:50.182736Z digest=sha256:38a2a1b9940d1a488465db927e56c4f763a9b2d73f90063fd914b13224a08b67

Observation 3ae028ca-7d4f-4477-91b8-157b9104151d · outbound

This paper cites Improving Large Language Model Fine-tuning for Solving Math Problems.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Improving Large Language Model Fine-tuning for Solving Math Problems

Reference 45

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:50.334485Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:50.334485Z digest=sha256:ac9a5254bb297cfb64bbc5a5bb4b6b7ba905683ca454e4a36cbf11541e65624e

Observation 3531bf4d-9dac-43e8-9824-003672414d38 · outbound

This paper cites Decoupled Weight Decay Regularization.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Decoupled Weight Decay Regularization

Reference 46

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:50.449308Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:50.449308Z digest=sha256:e8fecaecebd5223907cbf626cbbaa6ec04564315d27a921e30b2b377ac0d540d

Observation 9354b131-9533-4e5d-b039-e6df4926a553 · outbound

This paper cites Sgdr: Stochastic gradient descent with warm restarts.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Sgdr: Stochastic gradient descent with warm restarts

Reference 47

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.510379Z

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.

source=arxiv_source observed=2026-08-06T10:23:50.592820Z digest=sha256:eeb8012ebe3d390c6edc8004efab174b9f955eb9db1ca4cac9c6c0da3cec1d5e

Observation 4c731e40-810a-4bef-a614-2d93dff90ae5 · outbound

This paper cites On the sdes and scaling rules for adaptive gradient algorithms.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes On the sdes and scaling rules for adaptive gradient algorithms

Reference 48

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.500602Z

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.

source=arxiv_source observed=2026-08-06T10:23:50.740473Z digest=sha256:1aaa33ad45765eb4543c879bfce62c85c6ceda7a3c92efdcb3ee561dfbccf8fd

Observation 8218bda6-13f4-49a8-b80a-f68c516c1237 · outbound

This paper cites A kernel-based view of language model fine-tuning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes A kernel-based view of language model fine-tuning

Reference 49

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.490518Z

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.

source=arxiv_source observed=2026-08-06T10:23:50.856525Z digest=sha256:0b3902050e52cb9d1733130ae2835d9bf594e0f59463c665b00bc7c3fefc589b

Observation 003eaaf9-a0fc-42a3-bf0e-6e5dfb9aa8d6 · outbound

This paper cites Continuous-time limit of stochastic gradient descent revisited.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Continuous-time limit of stochastic gradient descent revisited

Reference 50

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.480815Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.026324Z digest=sha256:1fdbd97db32e39d276e805c43b4e46631ada5c9e24a6b8fcbf307a12c5dcac60

Observation 6d91ba18-c107-4815-8ff2-1f4b5fa364c7 · outbound

This paper cites Revisiting Small Batch Training for Deep Neural Networks.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Revisiting Small Batch Training for Deep Neural Networks

Reference 51

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.098062Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.098062Z digest=sha256:ec476cd2f7f520812840e3af89d2caf1356275d47ad1369fb7b26b8d794818b8

Observation 0c8c19e7-d57e-4c68-b2db-ed8c2d61560a · outbound

This paper cites Scaling data-constrained language models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Scaling data-constrained language models

Reference 52

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.471882Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.144262Z digest=sha256:8b406064a355dc587ec93102eb8f788c81b745a0b27a48b032b1d483cae074a7

Observation 20f73c62-3c1d-4e41-b91a-35c3b02ad8bf · outbound

This paper cites Smoothing data with faster moving averages.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Smoothing data with faster moving averages

Reference 53

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.461579Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.198764Z digest=sha256:3101f2258d1e89a4bc01a734cf813099ac17398b5ef02d26648b8f2c3a20cca9

Observation bc56b3ed-8cf8-43b1-a2bb-ef7d4ee063bc · outbound

This paper cites Introductory lectures on convex optimization: A basic course, volume 87.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Introductory lectures on convex optimization: A basic course, volume 87

Reference 54

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.248593Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.248593Z digest=sha256:e4f4c6bf6ba578d13834004b13fc4f97815be073c5de21eb911381255e2e0039

Observation 2c4a8d23-c44c-4622-a190-d8c2f52971b7 · outbound

This paper cites The AdEMAMix Optimizer: Better, Faster, Older.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes The AdEMAMix Optimizer: Better, Faster, Older

Reference 55

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.290978Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.290978Z digest=sha256:09fcb468926ac67257f5ea84744267f536b8ade477fdfed9104fd49b33f6a95b

Observation 903ef485-a89c-4c09-a73f-5d0fed6b6d38 · outbound

This paper cites Acceleration of stochastic approximation by averaging.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Acceleration of stochastic approximation by averaging

Reference 56

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.355993Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.355993Z digest=sha256:62914b35a72473feefbf42f234d6df4fa03e026ca3efa20ab4854fd5428d30a0

Observation f677e7bf-abe4-4933-aff1-388584f22d12 · outbound

This paper cites Early stopping-but when? In Neural Networks: Tricks of the trade, pages 55--69.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Early stopping-but when? In Neural Networks: Tricks of the trade, pages 55--69

Reference 57

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.442277Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.405126Z digest=sha256:cde159bdc15c3e7b718e5ce293af1c0736b0e0886e97f0ce99707e96fe9a38bb

Observation 2be369ba-6019-4b73-b08c-5dbffc90b2f0 · outbound

This paper cites Non-convex learning via stochastic gradient langevin dynamics: a nonasymptotic analysis.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Non-convex learning via stochastic gradient langevin dynamics: a nonasymptotic analysis

Reference 58

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.433596Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.485496Z digest=sha256:ee568bd380a1b229f0508246a9fa938840b662b9192e3451508c1d218ef18672

Observation 0701a113-76b3-454d-b5fa-9eb10308df36 · outbound

This paper cites Breaking the data barrier: a review of deep learning techniques for democratizing ai with small datasets.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Breaking the data barrier: a review of deep learning techniques for democratizing ai with small datasets

Reference 59

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.424548Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.560773Z digest=sha256:da354f90a9fe88d2f1a490694abcbc0d069c12a1e12d32e9c64723bea01c426e

Observation 2992878c-c911-417e-820b-29eedb5144d3 · outbound

This paper cites A stochastic approximation method.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes A stochastic approximation method

Reference 60

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.615896Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.615896Z digest=sha256:7226dd26a13deb299124f776a9f6dd7291385272dd1f2fb310793e63c66f875b

Observation 78351fb6-0bc2-4d51-b568-5f4c6915fac4 · outbound

This paper cites Computational-Statistical Tradeoffs at the Next-Token Prediction Barrier: Autoregressive and Imitation Learning under Misspecification.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Computational-Statistical Tradeoffs at the Next-Token Prediction Barrier: Autoregressive and Imitation Learning under Misspecification

Reference 61

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.683656Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.683656Z digest=sha256:b7a8ccd892b6de31a5043ec4bc3e739ecd4308df17b38397655ef4fa94844982

Observation aa313c8a-d8d0-4e06-a8a6-aee62b6b93d6 · outbound

This paper cites Efficient reductions for imitation learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Efficient reductions for imitation learning

Reference 62

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.748205Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.748205Z digest=sha256:488635e66c61fdcff9fd4a5a38a4b48b5a80b0445632b62c8714ba5ca9017c76

Observation b6c21b4a-dfc1-47d9-af30-0012d76567e7 · outbound

This paper cites A reduction of imitation learning and structured prediction to no-regret online learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes A reduction of imitation learning and structured prediction to no-regret online learning

Reference 63

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.794840Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.794840Z digest=sha256:2d57a81f32428173e4ff9a4feca235cb1aa8cc33a22fdce87bb9f8613d5f188a

Observation c787143b-5a90-431d-8842-6c8a82b87ebf · outbound

This paper cites Efficient estimations from a slowly convergent robbins-monro process.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Efficient estimations from a slowly convergent robbins-monro process

Reference 64

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.859645Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.859645Z digest=sha256:fd9420b8cfc864313c2b26f9b47671d3cf28fe9e906197e1dc2d39e1dbac99ad

Observation 8a87a096-54f8-4fa8-9708-00f1012e76a9 · outbound

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

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Training trajectories, mini-batch losses and the curious role of the learning rate

Reference 65

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.923341Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.923341Z digest=sha256:974b69655c8f4bb8ac1df4606d25f1683a994b7dcc6faa59ce8e92dd0c5fd721

Observation cbe8cf81-2aef-45e3-8691-11425c434cd4 · outbound

This paper cites Minimizing finite sums with the stochastic average gradient.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Minimizing finite sums with the stochastic average gradient

Reference 66

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.393897Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.987485Z digest=sha256:6c09691df8c2ed8b5df39687a3fc784fc40beae83d94cbde9b5987c6665d2c7e

Observation f8128de6-fac8-49a3-8b6f-0d8fac1217cd · outbound

This paper cites DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

Reference 67

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.990831Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.990831Z digest=sha256:c391a6d23cd44fac76f6baca6c68c15ffc81d464647f6e63cc727ab0ebdd6119

Observation 0aba7616-5249-4866-8dd2-021dbb866380 · outbound

This paper cites Super-convergence: Very fast training of neural networks using large learning rates.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Super-convergence: Very fast training of neural networks using large learning rates

Reference 68

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.384775Z

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.

source=arxiv_source observed=2026-08-06T10:23:51.993904Z digest=sha256:1e2da534371c1f7375019e91dc184519fe7652f2f6cc6e62a06978063bcb3679

Observation 543628a2-0b7b-4f9d-9b0b-1d21292efd3a · outbound

This paper cites Qwen2 Technical Report.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Qwen2 Technical Report

Reference 69

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:51.996778Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:51.996778Z digest=sha256:99b99304ebab409a0a6e4087687863f4676ced5979489428b25db721e9c81b72

Observation 38f0b125-781e-485d-97f1-451cb290f238 · outbound

This paper cites The calculus of variations.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes The calculus of variations

Reference 70

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.376386Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.000058Z digest=sha256:424380789556b19085eec7c6a7413bf1407b099afee7905adfd12db96a4195ed

Observation 49c508b4-ce11-4e0d-991f-627df44efe78 · outbound

This paper cites Position: Will we run out of data? limits of llm scaling based on human-generated data.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Position: Will we run out of data? limits of llm scaling based on human-generated data

Reference 71

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.367049Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.002693Z digest=sha256:b147ac594b4d65369d6729cff4a63892e3ac514d11f7845a17b609571a45c03f

Observation 06956e02-6b0b-4a44-a0a3-53510cf559dd · outbound

This paper cites Trl: Transformer reinforcement learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Trl: Transformer reinforcement learning

Reference 72

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.005485Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.005485Z digest=sha256:492feace191b918e36974d945935bdb53a17c36ba9c846db132c5bc45243cee7

Observation 5546061b-d571-40d7-bca1-0c66a26c29b5 · outbound

This paper cites SOAP: Improving and Stabilizing Shampoo using Adam.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes SOAP: Improving and Stabilizing Shampoo using Adam

Reference 73

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.008343Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.008343Z digest=sha256:1e757f05a3528712751b5fc87bb85eb05a8a3fc014e07c4ecb7470418964d263

Observation 49120fe7-4a2a-405e-88dd-11a465f3e2d1 · outbound

This paper cites Superglue: A stickier benchmark for general-purpose language understanding systems.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Superglue: A stickier benchmark for general-purpose language understanding systems

Reference 74

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.011334Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.011334Z digest=sha256:90361480f9cf9f38041a36cc35f5d8a2c3ec4ffcad888017de7452a6fb5e355e

Observation 58e3edb6-777e-47a0-b3a9-3d29950c1ede · outbound

This paper cites ema-pytorch: A simple way to keep track of an exponential moving average (ema) version of your pytorch model.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes ema-pytorch: A simple way to keep track of an exponential moving average (ema) version of your pytorch model

Reference 75

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.347595Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.014564Z digest=sha256:b86af087cb575597b8410388788da1307dbcbe674c9a140512452011c63a634a

Observation 62e13964-feeb-487e-83ed-eef21cfeb3f8 · outbound

This paper cites Chain-of-thought prompting elicits reasoning in large language models.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Chain-of-thought prompting elicits reasoning in large language models

Reference 76

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.017038Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.017038Z digest=sha256:f2bb2567c96c4587481571b498ded0b342e48adde1db3d72e06955cfd4a53b36

Observation f8442a29-4453-4dc7-8b4c-e7fde02af4a7 · outbound

This paper cites The large-sample distribution of the likelihood ratio for testing composite hypotheses.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes The large-sample distribution of the likelihood ratio for testing composite hypotheses

Reference 77

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.332725Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.019789Z digest=sha256:ed1aae059319602b6e03b89bd740f50192a01245362ec7a28ff3dc7a75da41c0

Observation be69d861-2136-41e0-9408-812959c91b77 · outbound

This paper cites HuggingFace's Transformers: State-of-the-art Natural Language Processing.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes HuggingFace's Transformers: State-of-the-art Natural Language Processing

Reference 78

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.022460Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.022460Z digest=sha256:2287c5f57a77da36331129c6e2286e0ca3fa73b0818b75ba401f6ad090d22d39

Observation a1d1c380-9e97-4495-84a3-df5afbb5e8af · outbound

This paper cites On early stopping in gradient descent learning.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes On early stopping in gradient descent learning

Reference 79

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.323689Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.025357Z digest=sha256:41c625d0d4442a4e564a3b609afd66e016f64bf5541342a008234accbe58b2fc

Observation cef1f308-1976-48a3-8571-c6680040d92a · outbound

This paper cites Which algorithmic choices matter at which batch sizes? insights from a noisy quadratic model.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Which algorithmic choices matter at which batch sizes? insights from a noisy quadratic model

Reference 80

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.314284Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.028226Z digest=sha256:29c85c493b0113bb34cdebff191c937692d820b26a5159cf93d43a79392fe492

Observation 31cb124b-8863-4e1e-92e1-f0b93d1a78dd · outbound

This paper cites How Does Critical Batch Size Scale in Pre-training?.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes How Does Critical Batch Size Scale in Pre-training?

Reference 81

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.031002Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.031002Z digest=sha256:80ee3c67f59f0bcff5af17c02afa5393d578ebe6663756c1908f5ea40ecd8eb6

Observation 18f89d3e-88e6-4ce6-b4be-7b8479fc860d · outbound

This paper cites Parameter identification for fractional ornstein--uhlenbeck processes based on discrete observation.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes Parameter identification for fractional ornstein--uhlenbeck processes based on discrete observation

Reference 82

Resolution
verified fuzzy
raw_fallback, observed 2026-08-06T10:23:52.304477Z

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.

source=arxiv_source observed=2026-08-06T10:23:52.034082Z digest=sha256:169d86ec7dd736c214d8429728f64c556c464cf10d7d30f5fba7873151abf341

Observation 203f65ac-9dcb-4883-b707-fd0aed7e2406 · outbound

This paper cites write newline.

EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes write newline

Reference 83

Resolution
unresolved
no resolver link, observed 2026-08-06T10:23:52.037279Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

source=arxiv_source observed=2026-08-06T10:23:52.037279Z digest=sha256:290e002188e4fe03b8abb2b7177b4907635a069756346ed16f5785505bc379f6

Pith citing papers

Observation 374ef32c-908d-4e31-9576-f1af93cdc24d · inbound

Training for the Model You Return: Improving Optimization for Iterate-Averaged Language Models cites this paper.

Training for the Model You Return: Improving Optimization for Iterate-Averaged Language Models EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes

Reference 3

Resolution
verified exact
arxiv_id, observed 2026-07-04T16:49:58.439604Z

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.

source=pdf_text observed=2026-06-26T00:09:54.353782Z digest=sha256:fa7d79cd7d296f31a6c38712fa356284b0b9c5bb41a280deeac2c40dd36857a7

Observation a0734cca-aa0f-4edd-a9cf-aafd1e765fd1 · inbound

Training for the Model You Return: Improving Optimization for Iterate-Averaged Language Models cites this paper.

Training for the Model You Return: Improving Optimization for Iterate-Averaged Language Models EMA Without the Lag: Bias-Corrected Iterate Averaging Schemes

Reference 3

Resolution
verified exact
arxiv_id, observed 2026-07-01T09:25:41.338906Z

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.

source=pdf_text observed=2026-07-01T06:36:46.924135Z digest=sha256:05196e34528031d8c447bc98d93f7a9d486b26d3348e83cd63367915838504a5