{"id":9,"date":"2022-02-04T14:31:08","date_gmt":"2022-02-04T14:31:08","guid":{"rendered":"https:\/\/dheerajnagaraj.com\/?page_id=9"},"modified":"2026-02-19T07:00:24","modified_gmt":"2026-02-19T07:00:24","slug":"publications","status":"publish","type":"page","link":"https:\/\/dheerajnagaraj.com\/index.php\/publications\/","title":{"rendered":"Publications"},"content":{"rendered":"\n<h3>Manuscripts:<\/h3>\n\n\n\n<ul><li><a href=\"https:\/\/arxiv.org\/abs\/2502.13450\" target=\"_blank\" rel=\"noreferrer noopener\" title=\"Interleaved Gibbs Diffusion for Constrained Generation\">Interleaved Gibbs Diffusion for Constrained Generation<\/a> (2025)<br>arXiv preprint<br>with Gautham Govind Anil, Sachin Yadav, Karthikeyan Shanmugam and Prateek Jain<\/li><\/ul>\n\n\n\n<h3 id=\"learning-from-markovian-trajectories\">Conference Publications:<\/h3>\n\n\n\n<div class=\"wp-container-1 wp-block-group\"><div class=\"wp-block-group__inner-container\">\n<ul id=\"block-2dd58cdc-ecdd-4548-96ab-cde43a7c8b2a\"><li><a href=\"https:\/\/arxiv.org\/abs\/2510.02692\">Fine-Tuning Diffusion Models via Intermediate Distribution Shaping<\/a> (2025)<br>ICLR (2026), to appear<br>with Gautham Govind Anil, Shaan Ul Haque, Nithish Kannen, Karthikeyan Shanmugam and Sanjay Shakkottai<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2506.07614\">Poisson Midpoint Method for Log Concave Sampling: Beyond the Strong Error Lower Bounds<\/a> (2025)<br>ICLR (2026), to appear<br>with Rishikesh Srinivasan<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2502.10354\" target=\"_blank\" rel=\"noreferrer noopener\">Dimension-free Score Matching and Time Bootstrapping for Diffusion Models<\/a> (2025)<br>NeurIPS (2025)<br>with Syamantak Kumar and Puranmrita Sarkar<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2503.13115\" target=\"_blank\" rel=\"noreferrer noopener\" title=\"Beyond Propagation of Chaos: A Stochastic Algorithm for Mean Field Optimization\">Beyond Propagation of Chaos: A Stochastic Algorithm for Mean Field Optimization<\/a> (2025)<br>COLT 2025 (to appear)<br>with Anant Raj and Chandan Tankala<\/li><li><a href=\"https:\/\/www.arxiv.org\/abs\/2405.17035\" target=\"_blank\" rel=\"noreferrer noopener\">Glauber Generative Model: Discrete Diffusion Models via Binary Classification<\/a> (2024)<br>ICLR 2025 <br>with Harshit Varma and Karthikeyan Shanmugam<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2405.17068\" target=\"_blank\" rel=\"noreferrer noopener\">The Poisson Midpoint Method for Langevin Dynamics: Provably Efficient Discretization for Diffusion Models<\/a> (2024)<br>NeurIPS 2024<br>with Saravanan Kandasamy<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2410.20135\" title=\"Near-Optimal Streaming Heavy-Tailed Statistical Estimation with Clipped SGD\u2028\">Near-Optimal Streaming Heavy-Tailed Statistical Estimation with Clipped SGD<\/a> (2024)<br>NeurIPS 2024<br>with Aniket Das, Soumyabrata Pal, Arun Suggala, Prateek Varshney<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2402.14807\">A Decision-Language Model (DLM) for Dynamic Restless Multi-Armed Bandit Tasks in Public Health <\/a>(2024)<br>NeurIPS 2024<br>with Nikhil Behari, Edwin Zhang, Yunfan Zhao, Aparna Taneja and Milind Tambe<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2306.09222\">Stochastic re-weighted gradient descent via distributionally robust optimization<\/a> (2023)<br>TMLR (to appear, 2024+)<br>with Ramnath Kumar, Kushal Majmundar and Arun Suggala<\/li><li><a href=\"https:\/\/arxiv.org\/pdf\/2310.14526.pdf\" title=\"Towards Zero Shot Learning in Restless Multi-armed Bandits\">Towards Zero Shot Learning in Restless Multi-armed Bandits<\/a> (2023)<br>IJCAI 2024<br>with Yunfan Zhao<em>,<\/em> Nikhil Behari, Edward Hughes, Edwin Zhang, Karl Tuyls, Aparna Taneja and Milind Tambe<\/li><\/ul>\n\n\n\n<ul id=\"block-2dd58cdc-ecdd-4548-96ab-cde43a7c8b2a\"><li><a href=\"https:\/\/arxiv.org\/abs\/2206.03171\" target=\"_blank\" rel=\"noreferrer noopener\">Look Back When Surprised: Stabilizing Reverse Experience Replay for Neural Approximation<\/a> (2022)<br>TMLR 2024<br><a href=\"http:\/\/arxiv.org\/abs\/2206.03171\" target=\"_blank\" rel=\"noreferrer noopener\" title=\"arXiv Preprint\">arXiv preprint<\/a><br>with Ramnath Kumar<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2305.17558\">Provably Fast Finite Particle Variants of SVGD via Virtual Particle Stochastic Approximation<\/a> (2023)<br>NeurIPS 2023 [Spotlight]<br>with Aniket Das<\/li><li><a href=\"http:\/\/arxiv.org\/abs\/2206.03792\" target=\"_blank\" rel=\"noreferrer noopener\">Utilising the CLT Structure in Stochastic Gradient based Sampling : Improved Analysis and Faster Algorithms<\/a> <br>COLT 2023<br>with Aniket Das and Anant Raj<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2306.14288\" title=\"Near Optimal Heteroscedastic Regression with Symbiotic Learning\">Near Optimal Heteroscedastic Regression with Symbiotic Learning<\/a><br>COLT 2023 <br>with Dheeraj Baby, Aniket Das and Praneeth Netrapalli<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2210.05355\">Multi-User Reinforcement Learning with Low Rank Rewards<\/a><br>ICML 2023 <br>with Naman Agarwal, Prateek Jain, Suhas Kowshik and Praneeth Netrapalli<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2210.05918\">Finite time analysis of temporal difference learning with linear function approximation: Tail averaging and regularisation <\/a><br><em>AISTATS 2023<\/em><br>with Gandharv Patil, L A Prashanth and Doina Precup<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2211.00112\">Indexability is Not Enough for Whittle: Improved, Near-Optimal Algorithms for Restless Bandits<\/a> <br><em>AAMAS 2023<\/em><br>with Abheek Ghosh, Manish Jain and Milind Tambe<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2110.08440\" target=\"_blank\" rel=\"noreferrer noopener\">Online Target Q-learning with Reverse Experience Replay: Efficiently finding the Optimal Policy for Linear MDPs<\/a> <br><em>ICLR 2022<\/em><br>with Naman Agarwal, Syomantak Chaudhuri, Prateek Jain and Praneeth Netrapalli.<\/li><\/ul>\n\n\n\n<ul><li><a href=\"https:\/\/arxiv.org\/abs\/2105.11558\" target=\"_blank\" rel=\"noreferrer noopener\">Near-optimal Offline and Streaming Algorithms for Learning Non-Linear Dynamical Systems.<\/a> (2021)<br><em>NeurIPS 2021 (spotlight)<\/em><br>with Prateek Jain, Suhas Kowshik and Praneeth Netrapalli.<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2103.05896\" target=\"_blank\" rel=\"noreferrer noopener\">Streaming Linear System Identification with Reverse Experience Replay<\/a> (2021)<br><em>NeurIPS 2021<\/em><br>with Prateek Jain, Suhas Kowshik and Praneeth Netrapalli.<\/li><\/ul>\n\n\n\n<ul id=\"block-d639b3e8-f65f-4bae-8117-5e4f3e254a16\"><li><a href=\"https:\/\/arxiv.org\/abs\/2108.10573\" target=\"_blank\" rel=\"noreferrer noopener\">The staircase property: How hierarchical structure can guide deep learning.<\/a> (2021)<br><em>NeurIPS 2021<\/em><br>with Emmanuel Abbe, Enric Boix-Adsera, Matthew Brennan and Guy Bresler<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2009.14444\" target=\"_blank\" rel=\"noreferrer noopener\">A Law of Robustness for Two-Layers Neural Networks<\/a>. (2020)<br><em>COLT 2021&nbsp;<\/em><br>with S\u00e9bastien Bubeck and Yuanzhi Li<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/2006.04048\" target=\"_blank\" rel=\"noreferrer noopener\">Sharp Representation Theorems for ReLU Networks with Precise Dependence on Depth.<\/a> (2020)<br><em>NeurIPS 2020&nbsp;<\/em><br>with Guy Bresler.<\/li><\/ul>\n\n\n\n<ul id=\"block-75805165-4a23-403b-bee6-9c09d1c09dd7\"><li><a href=\"https:\/\/arxiv.org\/abs\/2006.08916\" target=\"_blank\" rel=\"noreferrer noopener\">Least Squares Regression with Markovian Data: Fundamental Limits and Algorithms.<\/a> (2020)<br><em>NeurIPS 2020 (spotlight)<\/em><br>with Guy Bresler, Prateek Jain, Praneeth Netrapalli, Xian Wu<\/li><\/ul>\n\n\n\n<ul id=\"block-d639b3e8-f65f-4bae-8117-5e4f3e254a16\"><li><a href=\"http:\/\/proceedings.mlr.press\/v125\/bresler20a.html\" target=\"_blank\" rel=\"noreferrer noopener\">A Corrective View of Neural Networks: Representation, Memorization and Learning.<\/a> (2020)<br><em>COLT 2020<br>arXiv preprint : <a href=\"https:\/\/arxiv.org\/abs\/2002.00274\">arxiv:2002.00274<\/a><br><\/em>with Guy Bresler.<\/li><li><a href=\"https:\/\/arxiv.org\/abs\/1903.01463\" target=\"_blank\" rel=\"noreferrer noopener\">SGD without Replacement: Sharper Rates for General Smooth Convex Functions<\/a>. (2019)<br><em>ICML 2019<\/em><br>with Praneeth Netrapalli and Prateek Jain<\/li><li><a href=\"http:\/\/proceedings.mlr.press\/v75\/bresler18a\/bresler18a.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">Optimal Single Sample Tests for Structured versus Unstructured Network Data<\/a>. (2018)<br><em>COLT 2018.<\/em><br><em>arXiv preprint : <a href=\"https:\/\/arxiv.org\/abs\/1802.06186\">arxiv:1802.06186<\/a>&nbsp;<\/em><br>with Guy Bresler<\/li><\/ul>\n<\/div><\/div>\n\n\n\n<h3 id=\"probability-and-statistics\">Journal<strong> <\/strong>Publications<strong>:<\/strong><\/h3>\n\n\n\n<ul id=\"block-13963402-f32f-4481-ba0e-de62b3fde90c\"><li><a href=\"https:\/\/projecteuclid.org\/journals\/annals-of-applied-probability\/volume-34\/issue-1A\/Metastable-mixing-of-Markov-chains--Efficiently-sampling-low-temperature\/10.1214\/23-AAP1971.short\" target=\"_blank\" rel=\"noreferrer noopener\">Metastable Mixing of Markov Chains: Efficiently Sampling Low Temperature Exponential Random Graphs<\/a> (2024)<br><em>Annals of Applied Probability<\/em> <br><a href=\"http:\/\/arxiv.org\/abs\/2206.03171\" target=\"_blank\" rel=\"noreferrer noopener\" title=\"arXiv preprint\">arXiv preprint<\/a><br>with Guy Bresler and Eshaan Nichani<\/li><\/ul>\n\n\n\n<ul><li><a href=\"https:\/\/arxiv.org\/abs\/1904.12443\" target=\"_blank\" rel=\"noreferrer noopener\">Making Last Iterate of Stochastic Gradient Descent Information Theoretically Optimal<\/a>. (2021)<br><em>SIAM Journal on Optimization (2021) , COLT 2019 (extended abstract)<\/em><br>with Praneeth Netrapalli and Prateek Jain<\/li><li><a href=\"https:\/\/link.springer.com\/article\/10.1007\/s00440-020-00998-3\" target=\"_blank\" rel=\"noreferrer noopener\">Phase Transition for Detecting Latent Geometry in Random Graphs.<\/a> (2020)<br><em>Probability Theory and Related Fields<br>arXiv preprint: <a href=\"https:\/\/arxiv.org\/abs\/1910.14167\">arxiv:1910.14167<\/a><br><\/em>with Matthew Brennan and Guy Bresler<\/li><li><a href=\"https:\/\/projecteuclid.org\/euclid.aoap\/1571385634\" target=\"_blank\" rel=\"noreferrer noopener\">Stein&#8217;s Method for Stationary Distributions of Markov Chains and Application to Ising Models<\/a>. (2019)<br><em>Annals of Applied Probability&nbsp;<br>arXiv preprint: <a href=\"https:\/\/arxiv.org\/abs\/1712.05743\">arxiv:1712.05743<\/a>,<\/em><br>with Guy Bresler<\/li><li>&nbsp;<a href=\"https:\/\/journals.aps.org\/pra\/abstract\/10.1103\/PhysRevA.91.062304\" target=\"_blank\" rel=\"noreferrer noopener\">Continuous limit of discrete quantum walks<\/a>. (2015)<br><em>Physical Review A, <a href=\"https:\/\/arxiv.org\/abs\/1501.06950\">arxiv:1501.06950<\/a><\/em><br>with Todd A. Brun<\/li><\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Manuscripts: Interleaved Gibbs Diffusion for Constrained Generation (2025)arXiv preprintwith Gautham Govind Anil, Sachin Yadav, Karthikeyan Shanmugam and Prateek Jain Conference Publications: Fine-Tuning Diffusion Models via Intermediate Distribution Shaping (2025)ICLR (2026), to appearwith Gautham Govind Anil, Shaan Ul Haque, Nithish Kannen, Karthikeyan Shanmugam and Sanjay Shakkottai Poisson Midpoint Method for Log Concave Sampling: Beyond the Strong &hellip;<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0},"_links":{"self":[{"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/pages\/9"}],"collection":[{"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/comments?post=9"}],"version-history":[{"count":41,"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/pages\/9\/revisions"}],"predecessor-version":[{"id":141,"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/pages\/9\/revisions\/141"}],"wp:attachment":[{"href":"https:\/\/dheerajnagaraj.com\/index.php\/wp-json\/wp\/v2\/media?parent=9"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}