bibliography.bib

@article{cfg,
  author = {Jonathan Ho, Tim Salimans},
  journal = {NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications},
  title = {Classifier-Free Diffusion Guidance},
  year = {2021}
}

@article{wang2024boosting,
  title={Boosting Diffusion Models with an Adaptive Momentum Sampler.},
  author={Wang, Xiyu and Dinh, Anh-Dung and Liu, Daochang and Xu, Chang},
  journal={IJCAI},
  year={2024}
}

@article{bradley2024classifier,
  title={Classifier-Free Guidance is a Predictor-Corrector},
  author={Bradley, Arwen and Nakkiran, Preetum},
  journal={Mathematics of Modern Machine Learning (M3L) Workshop at NeurIPS},
  year={2024}
}

@article{
chidambaram2024what,
  title={What does guidance do? A fine-grained analysis in a simple setting},
  author={Muthu Chidambaram and Khashayar Gatmiry and Sitan Chen and Holden Lee and Jianfeng Lu},
  journal={NeurIPS},
  year={2024},
}

@article{karras2024guiding,
  title={Guiding a diffusion model with a bad version of itself},
  author={Karras, Tero and Aittala, Miika and Kynk{\"a}{\"a}nniemi, Tuomas and Lehtinen, Jaakko and Aila, Timo and Laine, Samuli},
  journal={NeurIPS},
  year={2024}
}

@article{kynkaanniemi2024applying,
  title={Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models},
  author={Kynkäänniemi, Tuomas and Aittala, Miika and Karras, Tero and Laine, Samuli and Aila, Timo and Lehtinen, Jaakko},
  year={2024},
  journal={NeurIPS}
}


@article{li2024understanding,
  title={Understanding Generalizability of Diffusion Models Requires Rethinking the Hidden Gaussian Structure},
  author={Li, Xiang and Dai, Yixiang and Qu, Qing},
  journal={NeurIPS},
  year={2024}
}

@article{qian2024boosting,
  title={Boosting Diffusion Models with Moving Average Sampling in Frequency Domain},
  author={Qian, Yurui and Cai, Qi and Pan, Yingwei and Li, Yehao and Yao, Ting and Sun, Qibin and Mei, Tao},
  journal={CVPR},
  year={2024}
}

@article{NEURIPS2023_d0da30e3,
 author = {Raya, Gabriel and Ambrogioni, Luca},
 journal={NeurIPS},
 title = {Spontaneous symmetry breaking in generative diffusion models},
 year = {2023}
}

@article{
wang2024the,
  title={The Unreasonable Effectiveness of Gaussian Score Approximation for Diffusion Models and its Applications},
  author={Binxu Wang and John Vastola},
  journal={TMLR},

  year={2024},

}


@article{10.5555/3692070.3694254,
author = {Wu, Yuchen and Chen, Minshuo and Li, Zihao and Wang, Mengdi and Wei, Yuting},
title = {Theoretical insights for diffusion guidance: a case study for Gaussian mixture models},
year = {2024},
journal = {ICML},
}

@article{zheng2023characteristic,
  title={Characteristic guidance: Non-linear correction for diffusion model at large guidance scale},
  author={Zheng, Candi and Lan, Yuan},
  journal={ICML},
  year={2023}
}

@article{de2025distributional,
  title={Distributional diffusion models with scoring rules},
  author={De Bortoli, Valentin and Galashov, Alexandre and Guntupalli, J Swaroop and Zhou, Guangyao and Murphy, Kevin and Gretton, Arthur and Doucet, Arnaud},
  journal={ICML},
  year={2025}
}

@article{gao2025reg,
  title={REG: Rectified Gradient Guidance for Conditional Diffusion Models},
  author={Gao, Zhengqi and Zha, Kaiwen and Zhang, Tianyuan and Xue, Zihui and Boning, Duane S},
  journal={ICML},
  year={2025}
}

@article{wang2024analysis,
  title={Analysis of classifier-free guidance weight schedulers},
  author={Wang, Xi and Dufour, Nicolas and Andreou, Nefeli and Cani, Marie-Paule and Abrevaya, Victoria Fern{\'a}ndez and Picard, David and Kalogeiton, Vicky},
  journal={TMLR},
  year={2024}
}

@article{Bansal_2023_CVPR,
    author    = {Bansal, Arpit and Chu, Hong-Min and Schwarzschild, Avi and Sengupta, Soumyadip and Goldblum, Micah and Geiping, Jonas and Goldstein, Tom},
    title     = {Universal Guidance for Diffusion Models},
    journal = {CVPR Workshops},
    year      = {2023},
}

@article{pavasovic2025classifierfreeguidancehighdimensionalanalysis,
      title={Classifier-Free Guidance: From High-Dimensional Analysis to Generalized Guidance Forms}, 
      author={Krunoslav Lehman Pavasovic and Jakob Verbeek and Giulio Biroli and Marc Mezard},
      year={2025},
      journal={arXiv},
}

@article{sadat2024eliminating,
  title={Eliminating oversaturation and artifacts of high guidance scales in diffusion models},
  author={Sadat, Seyedmorteza and Hilliges, Otmar and Weber, Romann M},
  journal={ICLR},
  year={2025}
}

@article{sadat2024no,
  title={No training, no problem: Rethinking classifier-free guidance for diffusion models},
  author={Sadat, Seyedmorteza and Kansy, Manuel and Hilliges, Otmar and Weber, Romann M},
  journal={ICLR},
  year={2025}
}

@article{shao2025images,
  title={Images Speak Louder Than Scores: Failure Mode Escape for Enhancing Generative Quality},
  author={Shao, Jie and Zhu, Ke and Fu, Minghao and Wang, Guo-hua and Wu, Jianxin},
  journal={arXiv:2508.09598},
  year={2025}
}

@article{skreta2025feynman,
  title={Feynman-kac correctors in diffusion: Annealing, guidance, and product of experts},
  author={Skreta, Marta and Akhound-Sadegh, Tara and Ohanesian, Viktor and Bondesan, Roberto and Aspuru-Guzik, Al{\'a}n and Doucet, Arnaud and Brekelmans, Rob and Tong, Alexander and Neklyudov, Kirill},
  journal={ICML},
  year={2025}
}

@article{skreta2024superposition,
  title={The Superposition of Diffusion Models Using the It$\backslash$\^{} o Density Estimator},
  author={Skreta, Marta and Atanackovic, Lazar and Bose, Avishek Joey and Tong, Alexander and Neklyudov, Kirill},
  journal={ICLR},
  year={2025}
}

@article{sun2025unified,
  title={Unified Continuous Generative Models},
  author={Sun, Peng and Jiang, Yi and Lin, Tao},
  journal={arXiv:2505.07447},
  year={2025}
}

@article{sun2025noise,
  title={Is Noise Conditioning Necessary for Denoising Generative Models?},
  author={Sun, Qiao and Jiang, Zhicheng and Zhao, Hanhong and He, Kaiming},
  journal={arXiv:2502.13129},
  year={2025}
}

@article{tang2025diffusion,
  title={Diffusion models without classifier-free guidance},
  author={Tang, Zhicong and Bao, Jianmin and Chen, Dong and Guo, Baining},
  journal={arXiv:2502.12154},
  year={2025}
}

@article{yang2025efficient,
  title={Efficient Training-Free High-Resolution Synthesis with Energy Rectification in Diffusion Models},
  author={Yang, Zhen and Shen, Guibao and Li, Minyang and Hou, Liang and Liu, Mushui and Wang, Luozhou and Tao, Xin and Wan, Pengfei and Zhang, Di and Chen, Ying-Cong},
  journal={arXiv:2503.02537},
  year={2025}
}

@article{zhao2025studying,
  title={Studying Classifier (-Free) Guidance From a Classifier-Centric Perspective},
  author={Zhao, Xiaoming and Schwing, Alexander G},
  journal={arXiv:2503.10638},
  year={2025}
}


@article{
sadat2024cads,
  title={{CADS}: Unleashing the Diversity of Diffusion Models through Condition-Annealed Sampling},
  author={Seyedmorteza Sadat and Jakob Buhmann and Derek Bradley and Otmar Hilliges and Romann M. Weber},
  journal={ICLR},
  year={2024},
  }


@article{chen2025s2guidancestochasticselfguidance,
      title={$S^2$-Guidance: Stochastic Self Guidance for Training-Free Enhancement of Diffusion Models}, 
      author={Chubin Chen and Jiashu Zhu and Xiaokun Feng and Nisha Huang and Meiqi Wu and Fangyuan Mao and Jiahong Wu and Xiangxiang Chu and Xiu Li},
      year={2025},
      journal={arXiv 2508.12880},
}

@article{shen2024understanding,
  title={Understanding and improving training-free loss-based diffusion guidance},
  author={Shen, Yifei and Jiang, Xinyang and Yang, Yifan and Wang, Yezhen and Han, Dongqi and Li, Dongsheng},
  journal={NeurIPS},
  year={2024}
}


@article{chen2024towards,
  title={Towards memorization-free diffusion models},
  author={Chen, Chen and Liu, Daochang and Xu, Chang},
  journal={CVPR},
  year={2024}
}

@article{karczewski2025devil,
  title={Devil is in the details: Density guidance for detail-aware generation with flow models},
  author={Karczewski, Rafa{\l} and Heinonen, Markus and Garg, Vikas},
  journal={ICML},
  year={2025}
}

@article{shah2025does,
  title={Does generation require memorization? creative diffusion models using ambient diffusion},
  author={Shah, Kulin and Kalavasis, Alkis and Klivans, Adam R and Daras, Giannis},
  journal={ICML},
  year={2025}
}


@article{dinh2023rethinking,
  title={Rethinking conditional diffusion sampling with progressive guidance},
  author={Dinh, Anh-Dung and Liu, Daochang and Xu, Chang},
  journal={NeurIPS},
  year={2023}
}

@article{zheng2022entropy,
  title={Entropy-driven sampling and training scheme for conditional diffusion generation},
  author={Zheng, Guangcong and Li, Shengming and Wang, Hui and Yao, Taiping and Chen, Yang and Ding, Shouhong and Li, Xi},
  journal={ECCV},
  year={2022},
}

@article{sehwag2022generating,
  title={Generating high fidelity data from low-density regions using diffusion models},
  author={Sehwag, Vikash and Hazirbas, Caner and Gordo, Albert and Ozgenel, Firat and Canton, Cristian},
  journal={CVPR},
  year={2022}
}


@article{karras2022elucidating,
  title={Elucidating the design space of diffusion-based generative models},
  author={Karras, Tero and Aittala, Miika and Aila, Timo and Laine, Samuli},
  journal={NeurIPS},
  year={2022}
}

@article{yu2023freedom,
  title={Freedom: Training-free energy-guided conditional diffusion model},
  author={Yu, Jiwen and Wang, Yinhuai and Zhao, Chen and Ghanem, Bernard and Zhang, Jian},
  journal={ICCV},
  year={2023}
}

@article{pmlr-v202-song23k,
  title = 	 {Loss-Guided Diffusion Models for Plug-and-Play Controllable Generation},
  author =       {Song, Jiaming and Zhang, Qinsheng and Yin, Hongxu and Mardani, Morteza and Liu, Ming-Yu and Kautz, Jan and Chen, Yongxin and Vahdat, Arash},
  year = 	 {2023},
  journal =    {PMLR},
}


# METRICS
@article{kynkaanniemi2019improved,
  title={Improved precision and recall metric for assessing generative models},
  author={Kynk{\"a}{\"a}nniemi, Tuomas and Karras, Tero and Laine, Samuli and Lehtinen, Jaakko and Aila, Timo},
  journal={NeurIPS},
  year={2019}
}

@article{sajjadi2018assessing,
  title={Assessing generative models via precision and recall},
  author={Sajjadi, Mehdi SM and Bachem, Olivier and Lucic, Mario and Bousquet, Olivier and Gelly, Sylvain},
  journal={NeurIPS},
  year={2018}
}

@article{heusel2017gans,
  title={Gans trained by a two time-scale update rule converge to a local nash equilibrium},
  author={Heusel, Martin and Ramsauer, Hubert and Unterthiner, Thomas and Nessler, Bernhard and Hochreiter, Sepp},
  journal={NeurIPS},
  year={2017}
}


@article{pmlr-v119-naeem20a,
  title = 	 {Reliable Fidelity and Diversity Metrics for Generative Models},
  author =       {Naeem, Muhammad Ferjad and Oh, Seong Joon and Uh, Youngjung and Choi, Yunjey and Yoo, Jaejun},
  journal = 	 {ICML},
  year = 	 {2020},
}

@article{salimans2016improved,
  title={Improved techniques for training gans},
  author={Salimans, Tim and Goodfellow, Ian and Zaremba, Wojciech and Cheung, Vicki and Radford, Alec and Chen, Xi},
  journal={NeurIPS},
  year={2016}
}

@article{theis2015note,
  title={A note on the evaluation of generative models},
  author={Theis, Lucas and Oord, A{\"a}ron van den and Bethge, Matthias},
  journal={ICLR},
  year={2016}
}

@article{kim2025diffusion,
  title={How Diffusion Models Memorize},
  author={Kim, Juyeop and Kim, Songkuk and Lee, Jong-Seok},
  journal={arXiv:2509.25705},
  year={2025}
}

@article{nalisnick2018deep,
  title={Do deep generative models know what they don't know?},
  author={Nalisnick, Eric and Matsukawa, Akihiro and Teh, Yee Whye and Gorur, Dilan and Lakshminarayanan, Balaji},
  journal={ICLR},
  year={2019}
}

@article{astolfi2024consistency,
  title={Consistency-diversity-realism Pareto fronts of conditional image generative models},
  author={Astolfi, Pietro and Careil, Marlene and Hall, Melissa and Ma{\~n}as, Oscar and Muckley, Matthew and Verbeek, Jakob and Soriano, Adriana Romero and Drozdzal, Michal},
  journal={arXiv:2406.10429},
  year={2024}
}


# Memorization

@article{song2025selectiveunderfittingdiffusionmodels,
      title={Selective Underfitting in Diffusion Models}, 
      author={Kiwhan Song and Jaeyeon Kim and Sitan Chen and Yilun Du and Sham Kakade and Vincent Sitzmann},
      year={2025},
      journal={arXiv:2510.01378},
}

@article{carlini2023extracting,
  title={Extracting training data from diffusion models},
  author={Carlini, Nicolas and Hayes, Jamie and Nasr, Milad and Jagielski, Matthew and Sehwag, Vikash and Tramer, Florian and Balle, Borja and Ippolito, Daphne and Wallace, Eric},
  journal={USENIX Security},
  year={2023}
}

@article{gu2025memorizationdiffusionmodels,
      title={On Memorization in Diffusion Models}, 
      author={Xiangming Gu and Chao Du and Tianyu Pang and Chongxuan Li and Min Lin and Ye Wang},
      year={2025},
      journal={TMLR}
}

@article{somepalli2023diffusion,
  title={Diffusion art or digital forgery? investigating data replication in diffusion models},
  author={Somepalli, Gowthami and Singla, Vasu and Goldblum, Micah and Geiping, Jonas and Goldstein, Tom},
  journal={CVPR},
  year={2023}
}

@article{van2021memorization,
  title={On memorization in probabilistic deep generative models},
  author={van den Burg, Gerrit and Williams, Chris},
  journal={NeurIPS},
  year={2021}
}

@article{yoon2023diffusion,
  title={Diffusion Probabilistic Models Generalize when They Fail to Memorize},
  author={TaeHo Yoon and Joo Young Choi and Sehyun Kwon and Ernest K. Ryu},
  journal={SPIGM Workshop - ICML},
  year={2023},
}

@article{wen2024detecting,
  title={Detecting, explaining, and mitigating memorization in diffusion models},
  author={Wen, Yuxin and Liu, Yuchen and Chen, Chen and Lyu, Lingjuan},
  journal={ICLR},
  year={2024}
}

@article{webster2023duplication,
  title={On the de-duplication of laion-2b},
  author={Webster, Ryan and Rabin, Julien and Simon, Loic and Jurie, Frederic},
  journal={arXiv:2303.12733},
  year={2023}
}

@article{buchanan2025edge,
  title={On the Edge of Memorization in Diffusion Models},
  author={Buchanan, Sam and Pai, Druv and Ma, Yi and De Bortoli, Valentin},
  journal={arXiv:2508.17689},
  year={2025}
}

@article{zhang2016understanding,
  title={Understanding deep learning requires rethinking generalization},
  author={Zhang, Chiyuan and Bengio, Samy and Hardt, Moritz and Recht, Benjamin and Vinyals, Oriol},
  journal={ICLR},
  year={2017}
}

@article{bartlett2020benign,
  title={Benign overfitting in linear regression},
  author={Bartlett, Peter L and Long, Philip M and Lugosi, G{\'a}bor and Tsigler, Alexander},
  journal={PNAS},
  year={2020},
}

@article{nakkiran2021deep,
  title={Deep double descent: Where bigger models and more data hurt},
  author={Nakkiran, Preetum and Kaplun, Gal and Bansal, Yamini and Yang, Tristan and Barak, Boaz and Sutskever, Ilya},
  journal={ICLR},
  year={2020},
}


@article{somepalli2023understanding,
  title={Understanding and mitigating copying in diffusion models},
  author={Somepalli, Gowthami and Singla, Vasu and Goldblum, Micah and Geiping, Jonas and Goldstein, Tom},
  journal={NeurIPS},
  year={2023}
}


@article{hintersdorf2024finding,
  title={Finding nemo: Localizing neurons responsible for memorization in diffusion models},
  author={Hintersdorf, Dominik and Struppek, Lukas and Kersting, Kristian and Dziedzic, Adam and Boenisch, Franziska},
  journal={NeurIPS},
  year={2024}
}

@article{zhang2024emergence,
  title={The emergence of reproducibility and consistency in diffusion models},
  author={Zhang, Huijie and Zhou, Jinfan and Lu, Yifu and Guo, Minzhe and Wang, Peng and Shen, Liyue and Qu, Qing},
  journal={ICML},
  year={2024}
}

@article{kamb2024analytic,
  title={An analytic theory of creativity in convolutional diffusion models},
  author={Kamb, Mason and Ganguli, Surya},
  journal={ICML},
  year={2025}
}

@article{feldman2020does,
  title={Does learning require memorization? a short tale about a long tail},
  author={Feldman, Vitaly},
  journal={ACM STOC},
  year={2020}
}


@article{NEURIPS2023_012af729,
	author = {Daras, Giannis and Shah, Kulin and Dagan, Yuval and Gollakota, Aravind and Dimakis, Alex and Klivans, Adam},
	journal = {NeurIPS},
	title = {Ambient Diffusion: Learning Clean Distributions from Corrupted Data},
	year = {2023},
  }