MS
Mrinank Sharma
cs.AIcs.LGcs.CLcs.CRCOVID-19epidemiologystat.MLcs.CVcs.SEGovernment
On Valency
published · living versionsW_3sjke4et·v1 · currentpublished
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
with Evan Hubinger, Carson Denison, Jesse Mu, Mike Lambert +34
1 version
Preprints & journals
32 papers in the corpus · 2019–2026Chain-of-Thought Hijacking2510.26418v4 · Jianli Zhao, Tingchen Fu, Rylan Schaeffer et al.2025 · 0 citationsarXiv
Who's in Charge? Disempowerment Patterns in Real-World LLM Usage2601.19062v1 · Mrinank Sharma, Miles McCain, Raymond Douglas et al.2026 · 2 citationsarXiv
Eliciting Harmful Capabilities by Fine-Tuning On Safeguarded Outputs2601.13528v1 · Jackson Kaunismaa, Avery Griffin, John Hughes et al.2026 · 0 citationsarXiv
Constitutional Classifiers++: Efficient Production-Grade Defenses against Universal Jailbreaks2601.04603v1 · Hoagy Cunningham, Jerry Wei, Zihan Wang et al.2026 · 0 citationsarXiv
PoisonBench: Assessing Large Language Model Vulnerability to Data Poisoning2410.08811v2 · Tingchen Fu, Mrinank Sharma, Philip Torr et al.2024 · 2 citationsarXiv
Changing composition of SARS-CoV-2 lineages and rise of Delta variant in England.34401689 · Mishra, Swapnil, Mindermann, Sören, Sharma, Mrinank et al.2025 · 130 citationsEClinicalMedicine. 2021;39:101064
Forecasting Rare Language Model Behaviors2502.16797v1 · Erik Jones, Meg Tong, Jesse Mu et al.2025 · 0 citationsarXiv
Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming2501.18837v1 · Mrinank Sharma, Meg Tong, Jesse Mu et al.2025 · 7 citationsarXiv
Best-of-N Jailbreaking2412.03556v2 · John Hughes, Sara Price, Aengus Lynch et al.2024 · 0 citationsarXiv
Failures to Find Transferable Image Jailbreaks Between Vision-Language Models2407.15211v2 · Rylan Schaeffer, Dan Valentine, Luke Bailey et al.2024 · 0 citationsarXiv
Jailbreak Defense in a Narrow Domain: Limitations of Existing Methods and a New Transcript-Classifier Approach2412.02159v1 · Tony T. Wang, John Hughes, Henry Sleight et al.2024 · 0 citationsarXiv
Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats2411.17693v1 · Jiaxin Wen, Vivek Hebbar, Caleb Larson et al.2024 · 1 citationarXiv
Rapid Response: Mitigating LLM Jailbreaks with a Few Examples2411.07494v1 · Alwin Peng, Julian Michael, Henry Sleight et al.2024 · 0 citationsarXiv
Incorporating Unlabelled Data into Bayesian Neural Networks2304.01762v3 · Mrinank Sharma, Tom Rainforth, Yee Whye Teh et al.2023 · 0 citationsarXiv
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training2401.05566v3 · Evan Hubinger, Carson Denison, Jesse Mu et al.2024 · 39 citationsarXivon Valency
Prioritized training on points that are learnable, worth learning, and not yet learned (workshop version)2107.02565v4 · Soren Mindermann, Muhammed Razzak, Winnie Xu et al.2021 · 0 citationsICML 2021 Workshop on Subset Selection in Machine Learning
Understanding and Controlling a Maze-Solving Policy Network2310.08043v1 · Ulisse Mini, Peli Grietzer, Mrinank Sharma et al.2023 · 0 citationsarXiv
Effectiveness assessment of non-pharmaceutical interventions: lessons learned from the COVID-19 pandemic.36965985 · Lison, Adrian, Banholzer, Nicolas, Sharma, Mrinank et al.2023 · 76 citationsThe Lancet. Public health. 2023;8(4):e311-e317
Do Bayesian Neural Networks Need To Be Fully Stochastic?2211.06291v2 · Mrinank Sharma, Sebastian Farquhar, Eric Nalisnick et al.2022 · 5 citationsarXiv
Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learnt2206.07137v3 · Soren Mindermann, Jan Brauner, Muhammed Razzak et al.2022 · 19 citationsarXiv
Seasonal variation in SARS-CoV-2 transmission in temperate climates: A Bayesian modelling study in 143 European regions.36026483 · Gavenčiak, Tomáš, Monrad, Joshua Teperowski, Leech, Gavin et al.2022 · 58 citationsPLoS computational biology. 2022;18(8):e1010435
Mask wearing in community settings reduces SARS-CoV-2 transmission.35639701 · Leech, Gavin, Rogers-Smith, Charlie, Monrad, Joshua Teperowski et al.2022 · 117 citationsProceedings of the National Academy of Sciences of the United States of America. 2022;119(23):e2119266119
A dataset of non-pharmaceutical interventions on SARS-CoV-2 in Europe.35365668 · Altman, George, Ahuja, Janvi, Monrad, Joshua Teperowski et al.2022 · 12 citationsScientific data. 2022;9(1):145
Understanding the effectiveness of government interventions against the resurgence of COVID-19 in Europe.34611158 · Sharma, Mrinank, Mindermann, Sören, Rogers-Smith, Charlie et al.2021 · 230 citationsNature communications. 2021;12(1):5820
Seasonal variation in SARS-CoV-2 transmission in temperate climates10.1101/2021.06.10.21258647v3 · Gavenciak, T., Monrad, J. T., Leech, G. et al.2021 · 27 citationsmedRxiv
Is the cure really worse than the disease? The health impacts of lockdowns during COVID-19.34281914 · Meyerowitz-Katz, Gideon, Bhatt, Samir, Ratmann, Oliver et al.2021 · 111 citationsBMJ global health. 2021;6(8)
Mass mask-wearing notably reduces COVID-19 transmission10.1101/2021.06.16.21258817v1 · Leech, G., Rogers-Smith, C., Sandbrink, J. B. et al.2021 · 12 citationsmedRxiv
How Robust are the Estimated Effects of Nonpharmaceutical Interventions against COVID-19?2007.13454v3 · Mrinank Sharma, Soren Mindermann, Jan Markus Brauner et al.2020 · 18 citationsNeurIPS 2020, Advances in Neural Information Processing Systems 33
Understanding the effectiveness of government interventions in Europe's second wave of COVID-1910.1101/2021.03.25.21254330v1 · Sharma, M., Mindermann, S., Rogers-Smith, C. et al.2021 · 45 citationsmedRxiv
Inferring the effectiveness of government interventions against COVID-19.33323424 · Brauner, Jan M, Mindermann, Sören, Sharma, Mrinank et al.2021 · 1,140 citationsScience (New York, N.Y.). 2021;371(6531)
The effectiveness of eight nonpharmaceutical interventions against COVID-19 in 41 countries10.1101/2020.05.28.20116129v4 · Brauner, J. M., Mindermann, S., Sharma, M. et al.2020 · 88 citationsmedRxiv
Differentially Private Federated Variational Inference1911.10563v1 · Mrinank Sharma, Michael Hutchinson, Siddharth Swaroop et al.2019 · 1 citationarXiv
Career total: 41 works. 32 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.