Paper Citation Record · LEDGER

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

As of 8 August 2026, this Paper Citation Record lists 57 of 57 outbound references and 7 inbound Pith citation observations for arXiv:2505.16792.

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

pith.paper-citation-record.v1
2505.16792 v1

Coverage vector

measured 57 of 57 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-08-07T15:00:41.559892Z

measured 64 of 64 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-08T06:32:00.761636+00:00

measured 7 of 7 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-08-03T21:07:15.008401Z

measured 0 of 1 external citation measurements

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

Source: arxiv_reference, observed 2026-07-03T14:38:28.535602Z

Reference resolution

57 of 57 outbound references displayed

verified exact1
verified fuzzy18
unresolved38
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 65e044ad-0280-4194-be71-3321094d0c30 · outbound

This paper cites All are worth words: A vit backbone for diffusion models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training All are worth words: A vit backbone for diffusion models

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:34.942442Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2838bf3d-b02e-4edf-8eaa-809c99848c97 · outbound

This paper cites Video generation models as world simulators.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Video generation models as world simulators

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:45.147720Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation ffa6450b-3a43-4dc6-b9eb-ee01306ed6e1 · outbound

This paper cites Rethinking attention with performers.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Rethinking attention with performers

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.103310Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 936b1f20-f73e-4bb4-b3ac-77beb37443b3 · outbound

This paper cites Imagenet: A large-scale hierarchical image database.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Imagenet: A large-scale hierarchical image database

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.197593Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c7574565-d664-486a-82ff-437bd3234793 · outbound

This paper cites Diffusion models beat gans on image synthesis.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Diffusion models beat gans on image synthesis

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.297995Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7aca5932-f3a6-46cb-bf7d-9ffadc24f0ff · outbound

This paper cites An image is worth 16x16 words: Transformers for image recognition at scale.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training An image is worth 16x16 words: Transformers for image recognition at scale

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.396068Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 309d6daa-dcc5-4bff-8e51-b0325f6cfca2 · outbound

This paper cites Sigmoid-weighted linear units for neural network function approximation in reinforcement learning.Neural Networks, 107:3–11, 2018.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Sigmoid-weighted linear units for neural network function approximation in reinforcement learning.Neural Networks, 107:3–11, 2018

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.448492Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 74eff953-e867-4957-a146-39ba4c923e6f · outbound

This paper cites Scaling rectified flow transformers for high-resolution image synthesis.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Scaling rectified flow transformers for high-resolution image synthesis

Reference 8

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.505927Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3c68f45c-5c8c-4a5f-b5d0-9c1a081a2749 · outbound

This paper cites Dit4edit: Diffusion transformer for image editing.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Dit4edit: Diffusion transformer for image editing

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.984595Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 99fca000-fee4-43fc-acbc-d999ebc7d7de · outbound

This paper cites Masked diffusion transformer is a strong image synthesizer.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Masked diffusion transformer is a strong image synthesizer

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.816956Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 9f270128-508b-4203-8517-6dbd5f9c8f3d · outbound

This paper cites MDTv2: Masked Diffusion Transformer is a Strong Image Synthesizer.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training MDTv2: Masked Diffusion Transformer is a Strong Image Synthesizer

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.728615Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bedc7bd7-ac65-4dd6-8ded-11a3d05f5dc7 · outbound

This paper cites Efficient diffusion training via min-snr weighting strategy.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Efficient diffusion training via min-snr weighting strategy

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.650391Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation ceaec345-079a-42d1-ab12-1968ac0ef075 · outbound

This paper cites Momentum contrast for unsupervised visual representation learning.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Momentum contrast for unsupervised visual representation learning

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:35.910309Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d997a8bf-9b02-425d-9e57-00fb4c21ad4a · outbound

This paper cites Masked autoencoders are scalable vision learners.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Masked autoencoders are scalable vision learners

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:36.087592Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6a3e052d-a6ec-448c-a543-59ba3d7abbf0 · outbound

This paper cites Gans trained by a two time-scale update rule converge to a local nash equilibrium.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Gans trained by a two time-scale update rule converge to a local nash equilibrium

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:36.244565Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8a1de0a7-596e-416c-a6b4-470fa080a851 · outbound

This paper cites Classifier-free diffusion guidance.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Classifier-free diffusion guidance

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:36.397671Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 20f8bc7e-380d-4ed4-a47a-4213845a5848 · outbound

This paper cites Denoising diffusion probabilistic models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Denoising diffusion probabilistic models

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.437079Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation c800f9e4-5645-4dc7-9d03-27cd286f8622 · outbound

This paper cites Video diffusion models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Video diffusion models

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:36.714017Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9bb0220d-01dd-4711-abb3-134ac9250aaa · outbound

This paper cites Do compressed llms forget knowledge? an experimental study with practical implications.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Do compressed llms forget knowledge? an experimental study with practical implications

Reference 20

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.291346Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 6867fc13-e941-49b9-a580-e1f8338c8e2b · outbound

This paper cites Analyzing and improving the training dynamics of diffusion models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Analyzing and improving the training dynamics of diffusion models

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:37.020554Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3b8f9b01-cb7a-4809-bc6d-4f24c6f3fb31 · outbound

This paper cites Transformers are rnns: fast autoregressive transformers with linear attention.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Transformers are rnns: fast autoregressive transformers with linear attention

Reference 22

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:37.153074Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 21f5931c-d5a0-4728-a6b1-e5e206bc7e44 · outbound

This paper cites Soft truncation: A universal training technique of score-based diffusion model for high precision score estimation.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Soft truncation: A universal training technique of score-based diffusion model for high precision score estimation

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:44.135669Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation b046cd0b-704b-4d79-88cb-3f9f7e863d26 · outbound

This paper cites Understanding diffusion objectives as the elbo with simple data augmentation.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Understanding diffusion objectives as the elbo with simple data augmentation

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.962159Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation ec0547b1-e00c-415c-897b-4188683d32cd · outbound

This paper cites Kingma and Jimmy Ba.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Kingma and Jimmy Ba

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:37.578491Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0dfc77de-50e2-4887-8c95-bd50ee6decae · outbound

This paper cites Tread: Token routing for efficient architecture-agnostic diffusion training.arXiv preprint arXiv:2501.04765, 2025.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Tread: Token routing for efficient architecture-agnostic diffusion training.arXiv preprint arXiv:2501.04765, 2025

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:37.700485Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8abed8a9-2e0d-4ac8-a928-ec69db307542 · outbound

This paper cites Improved precision and recall metric for assessing generative models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Improved precision and recall metric for assessing generative models

Reference 27

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:37.839953Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f68055b0-cb08-4b30-b48b-8b23a22741f3 · outbound

This paper cites Applying guidance in a limited interval improves sample and distribution quality in diffusion models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Applying guidance in a limited interval improves sample and distribution quality in diffusion models

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.808398Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 101922a0-4b4d-4728-b1e0-1d00fac9984b · outbound

This paper cites Flux.https://github.com/black-forest-labs/flux, 2024.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Flux.https://github.com/black-forest-labs/flux, 2024

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:38.037070Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d35d8f7a-8a21-4fdc-9761-15c559a3c030 · outbound

This paper cites On the surprising effectiveness of attention transfer for vision transformers.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training On the surprising effectiveness of attention transfer for vision transformers

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.644889Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation ea6e1416-9557-4cbc-bc35-8f5a4e9e7405 · outbound

This paper cites Lawrence Zitnick.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Lawrence Zitnick

Reference 32

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.479492Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 72598e8b-924b-474f-8849-de45e5e71a21 · outbound

This paper cites Decoupled weight decay regularization.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Decoupled weight decay regularization

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:38.804245Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6916ee39-6ee4-4587-9ac3-2b40d4af4381 · outbound

This paper cites SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers

Reference 34

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:38.912442Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bc446cda-ff38-4c71-84d4-c86790016c86 · outbound

This paper cites Generating images with sparse representations.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Generating images with sparse representations

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.334608Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 1d89bbd2-a64b-4e79-9a03-9f9599217c20 · outbound

This paper cites an unresolved cited work.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Unresolved cited work

Reference 36

Resolution: unresolved; raw_fallback, observed 2026-08-07T15:00:43.174818Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 58a37da5-c2aa-4b37-a12e-fa2085fc4028 · outbound

This paper cites Scalable diffusion models with transformers.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Scalable diffusion models with transformers

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:39.340945Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a8432fa5-df9e-434c-bdde-a437b7352ffb · outbound

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

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Learning transferable visual models from natural language supervision

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:39.494926Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7400ed3c-d70d-4907-8b63-e6e27434abb1 · outbound

This paper cites Tinymim: An empirical study of distilling mim pre-trained models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Tinymim: An empirical study of distilling mim pre-trained models

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:43.001117Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 9a3cce93-0a50-4b8c-a373-911b5a22257f · outbound

This paper cites High-resolution image synthesis with latent diffusion models.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training High-resolution image synthesis with latent diffusion models

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:39.829955Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1f23464f-2cbe-4924-ae54-73c8a590631d · outbound

This paper cites U-net: Convolutional networks for biomedical image segmentation.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training U-net: Convolutional networks for biomedical image segmentation

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:39.937291Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 21669526-a3c5-44cb-9a08-0c544ee572ce · outbound

This paper cites Improved techniques for training gans.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Improved techniques for training gans

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.019413Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a2107822-0998-4a9f-9971-5073714d67aa · outbound

This paper cites Deep unsupervised learning using nonequilibrium thermodynamics.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Deep unsupervised learning using nonequilibrium thermodynamics

Reference 43

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.118868Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 97b730a0-bd52-463c-9748-b26db7dfdc8b · outbound

This paper cites Score-based generative modeling through stochastic differential equations.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Score-based generative modeling through stochastic differential equations

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.214782Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c6087e4e-fb0d-4967-9b9a-afdcfbfd3ba4 · outbound

This paper cites U-repa: Aligning diffusion u-nets to vits.arXiv preprint arXiv:2503.18414, 2025.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training U-repa: Aligning diffusion u-nets to vits.arXiv preprint arXiv:2503.18414, 2025

Reference 45

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.287192Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ef7a139d-5334-433b-9dc4-a856a46f7a3a · outbound

This paper cites Attention is all you need.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Attention is all you need

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:42.831026Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 22926fc2-83b0-4a28-a288-9db85312e0a3 · outbound

This paper cites Lit: Delving into a simplified linear diffusion transformer for image generation.arXiv preprint arXiv:2501.12976, 2025.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Lit: Delving into a simplified linear diffusion transformer for image generation.arXiv preprint arXiv:2501.12976, 2025

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.518657Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 0c29e818-7619-4388-adc1-5936e06e1b0b · outbound

This paper cites Attention distillation: self-supervised vision transformer students need more guidance.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Attention distillation: self-supervised vision transformer students need more guidance

Reference 48

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:42.654925Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 1fe3d85b-6249-421c-a498-b1b5d6e49774 · outbound

This paper cites A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.693309Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c601da6e-7c88-4728-a88d-01899f032bd5 · outbound

This paper cites Linformer: Self-Attention with Linear Complexity.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Linformer: Self-Attention with Linear Complexity

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.775339Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e72b801c-7067-4cd1-90f8-49621894e3cd · outbound

This paper cites SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.850273Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9d686c22-5d27-4927-ae7f-a034ea1f9227 · outbound

This paper cites Reconstruction vs.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Reconstruction vs

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:41.054346Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2b98be7f-d3dd-4796-ad24-b1cb77e1a5ee · outbound

This paper cites SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:40.984107Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b4e40034-919d-426d-8390-d217aa00b346 · outbound

This paper cites Representation alignment for generation: Training diffusion transformers is easier than you think.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Representation alignment for generation: Training diffusion transformers is easier than you think

Reference 55

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:42.382237Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 0dfff8e2-5eca-43b5-87b8-a068a1ae4bfb · outbound

This paper cites Fasterdit: Towards faster diffusion transformers training without architecture modification.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Fasterdit: Towards faster diffusion transformers training without architecture modification

Reference 56

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:42.540341Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 4a1f0f0b-57fd-42ac-a407-af6d5ee50e89 · outbound

This paper cites Attention Distillation: A Unified Approach to Visual Characteristics Transfer.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Attention Distillation: A Unified Approach to Visual Characteristics Transfer

Reference 57

Resolution: verified exact; local_arxiv, observed 2026-08-07T15:00:41.731004Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation cc102fb3-24d9-4b51-9c7c-85987e89b8a3 · outbound

This paper cites Fast training of diffusion models with masked transformers.TMLR, 2024.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Fast training of diffusion models with masked transformers.TMLR, 2024

Reference 58

Resolution: verified fuzzy; raw_fallback, observed 2026-08-07T15:00:42.267057Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 5b9c3cd3-a861-4d28-b8c7-d445db19c7e4 · outbound

This paper cites DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention

Reference 2024

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:41.559892Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 77172420-d6fb-4830-bde2-cacb9db93f2f · outbound

This paper cites an unresolved cited work.

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training Unresolved cited work

Reference 2025

Resolution: unresolved; no resolver link, observed 2026-08-07T15:00:38.349204Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Pith citing papers

Observation 9bcd6b93-15a5-4b6f-8d60-31777eb09352 · inbound

Align & Invert: Solving Inverse Problems with Diffusion and Flow-based Models via Representation Alignment cites this paper.

Align & Invert: Solving Inverse Problems with Diffusion and Flow-based Models via Representation Alignment REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 2004

Resolution: unresolved; no resolver link, observed 2026-08-03T21:07:15.008401Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a8d9182a-c01b-4eed-b82d-4fb6cc27b3bb · inbound

Self-transcendence: Is External Feature Guidance Indispensable for Accelerating Diffusion Transformer Training? cites this paper.

Self-transcendence: Is External Feature Guidance Indispensable for Accelerating Diffusion Transformer Training? REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 46

Resolution: unresolved; no resolver link, observed 2026-08-03T11:04:33.326758Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 29efbc0f-9532-4114-bd04-26342bbf6e21 · inbound

Under One Sun: Multi-Object Generative Perception of Materials and Illumination cites this paper.

Under One Sun: Multi-Object Generative Perception of Materials and Illumination REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 47

Resolution: unresolved; no resolver link, observed 2026-07-13T22:08:47.493022Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4e8f97e0-eb6d-445f-b916-6f14052c13cc · inbound

TORA: Topological Representation Alignment for 3D Shape Assembly cites this paper.

TORA: Topological Representation Alignment for 3D Shape Assembly REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 50

Resolution: verified exact; arxiv_id, observed 2026-05-13T17:13:01.215375Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 1c0c920a-558a-4ff7-9cbb-366ddd544412 · inbound

Generate in Reconstruction Space, Match in Semantic Space: Transport Geometry for One-Step Generation cites this paper.

Generate in Reconstruction Space, Match in Semantic Space: Transport Geometry for One-Step Generation REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 44

Resolution: verified exact; arxiv_id, observed 2026-06-28T19:02:34.284232Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 5520c885-4ece-4346-9b85-6e294ff32670 · inbound

MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training cites this paper.

MaskAlign: Token-Subset Representation Alignment for Efficient Diffusion Training REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 26

Resolution: verified exact; arxiv_id, observed 2026-07-02T22:37:26.187428Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.

Observation 90edfcdd-c117-47e9-bb7c-b6afd71ad878 · inbound

From SRA to Self-Flow: Data Augmentation or Self-Supervision? cites this paper.

From SRA to Self-Flow: Data Augmentation or Self-Supervision? REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

Reference 37

Resolution: verified exact; arxiv_id, observed 2026-07-03T14:38:28.537286Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-08T06:32:00.761636+00:00.