Brian Christian — publicationsNew papers and preprints.https://brianchristian.org/Self-blinding and counterfactual self-simulation mitigate biases and sycophancy in large language modelshttps://arxiv.org/abs/2601.14553https://arxiv.org/abs/2601.14553Brian Christian, Matan MazorThu, 01 Jan 2026 00:00:00 GMTPost-training makes large language models less human-likehttps://arxiv.org/abs/2605.07632https://arxiv.org/abs/2605.07632Marcel Binz, Elif Akata, Abdullah Almaatouq, Brian Christian, Eric SchulzThu, 01 Jan 2026 00:00:00 GMTAI epistemic risks: Emerging mechanisms & evidencehttps://doi.org/10.2139/ssrn.6873005https://doi.org/10.2139/ssrn.6873005Mick Yang, Stephen Casper, Jonathan Stray, Jasmine Li, Cameron Jones, Anna Gausen, Natasha Jaques, Brian ChristianThu, 01 Jan 2026 00:00:00 GMTAI assistance reduces persistence and hurts independent performancehttps://arxiv.org/abs/2604.04721https://arxiv.org/abs/2604.04721Grace Liu, Brian Christian, Tsvetomira Dumbalska, Michiel Bakker, Rachit Dubey — Proceedings of the Conference on Language Modeling (COLM 2026)Thu, 01 Jan 2026 00:00:00 GMTResolving Feynman’s restaurant problem reveals optimal solutions and human strategieshttps://doi.org/10.1073/pnas.2509612123https://doi.org/10.1073/pnas.2509612123Brian Christian, Evan M. Russek, Thomas L. Griffiths — Proceedings of the National Academy of Sciences (PNAS)Thu, 01 Jan 2026 00:00:00 GMTReward models inherit value biases from pretraininghttps://openreview.net/forum?id=dT399j1Azvhttps://openreview.net/forum?id=dT399j1AzvBrian Christian, Jessica A.F. Thompson, Elle Michelle Yang, Vincent Adam, Hannah Rose Kirk, Christopher Summerfield, Tsvetomira Dumbalska — Proceedings of the Fourteenth International Conference on Learning Representations (ICLR 2026)Thu, 01 Jan 2026 00:00:00 GMTReward model interpretability via optimal and pessimal tokenshttps://brianchristian.org/research/papers/2025_Christian_etal_Reward_Model_Interpretability.pdfhttps://brianchristian.org/research/papers/2025_Christian_etal_Reward_Model_Interpretability.pdfBrian Christian, Hannah Rose Kirk, Jessica A.F. Thompson, Christopher Summerfield, Tsvetomira Dumbalska — Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency (FAccT 2025)Wed, 01 Jan 2025 00:00:00 GMTComputational frameworks for human carehttps://brianchristian.org/research/papers/2025_Christian_Computational_Frameworks_for_Human_Care.pdfhttps://brianchristian.org/research/papers/2025_Christian_Computational_Frameworks_for_Human_Care.pdfBrian Christian — Dædalus: The Social Science of CaregivingWed, 01 Jan 2025 00:00:00 GMTUsing adaptive intrinsic motivation in RL to model learning across developmenthttps://brianchristian.org/research/papers/2024_Sandbrink_etal_Using_Adaptive_Intrinsic_Motivation.pdfhttps://brianchristian.org/research/papers/2024_Sandbrink_etal_Using_Adaptive_Intrinsic_Motivation.pdfKai Sandbrink, Brian Christian, Linas Nasvytis, Christian Schroeder de Witt, Patrick Butlin — NeurIPS Workshop on Intrinsically-Motivated and Open-Ended Learning (IMOL)Mon, 01 Jan 2024 00:00:00 GMTPersonhood credentials: Artificial intelligence and the need for privacy-preserving ways to distinguish who is real onlinehttps://arxiv.org/abs/2408.07892https://arxiv.org/abs/2408.07892Steven Adler, Zoë Hitzig, Shrey Jain, Catherine Brewer, Wayne Chang, Renée DiResta, Brian ChristianMon, 01 Jan 2024 00:00:00 GMTCan reinforcement learning model learning across development? Online lifelong learning through adaptive intrinsic motivationhttps://brianchristian.org/research/papers/2024_Sandbrink_etal_RL_Model_Learning.pdfhttps://brianchristian.org/research/papers/2024_Sandbrink_etal_RL_Model_Learning.pdfKai Sandbrink, Brian Christian, Linas Nasvytis, Christian Schroeder de Witt, Patrick Butlin — Proceedings of the 46th Annual Conference of the Cognitive Science SocietyMon, 01 Jan 2024 00:00:00 GMTAI Objectives Institute whitepaper: A research agenda for the production of a flourishing civilizationhttps://ai.objectives.institute/s/AOI-Whitepaper.pdfhttps://ai.objectives.institute/s/AOI-Whitepaper.pdfPeter Eckersley, Max Shron, Tushant Jha, Divya Siddarth, Brittney Gallagher, Carroll Wainwright, Joel Lehman, Brian Christian, Deger TuranSun, 01 Jan 2023 00:00:00 GMTComment of the AI Policy and Governance Working Group on the NTIA AI Accountability Policy Request for Comment Docket NTIA-230407-0093https://www.ias.edu/sites/default/files/AI%20Policy%20and%20Governance%20Working%20Group%20NTIA%20Comment.pdfhttps://www.ias.edu/sites/default/files/AI%20Policy%20and%20Governance%20Working%20Group%20NTIA%20Comment.pdfAaron Maniam, Alondra Nelson, Ben Garfinkel, Brian Christian, Daniel E. Ho, Dorothy Chou, Helen Toner, Inioluwa Deborah RajiSun, 01 Jan 2023 00:00:00 GMTHow do humans overcome individual computational limitations by working together?https://doi.org/10.1111/cogs.13232https://doi.org/10.1111/cogs.13232Natalia Vélez, Brian Christian, Mathew Hardy, Bill D. Thompson, Thomas L. Griffiths — Cognitive ScienceSun, 01 Jan 2023 00:00:00 GMTCaching algorithms and rational models of memoryhttps://brianchristian.org/research/papers/2014_Press_etal_Caching_Algorithms.pdfhttps://brianchristian.org/research/papers/2014_Press_etal_Caching_Algorithms.pdfAvi Press, M Pacer, Thomas L. Griffiths, Brian Christian — Proceedings of the Annual Meeting of the Cognitive Science SocietyWed, 01 Jan 2014 00:00:00 GMTUsing category structures to test iterated learning as a method for identifying inductive biaseshttps://doi.org/10.1080/03640210701801974https://doi.org/10.1080/03640210701801974Thomas L. Griffiths, Brian Christian, Michael L. Kalish — Cognitive ScienceTue, 01 Jan 2008 00:00:00 GMTRevealing priors on category structures through iterated learninghttps://brianchristian.org/research/papers/2006_Griffiths_etal_Revealing_Priors.pdfhttps://brianchristian.org/research/papers/2006_Griffiths_etal_Revealing_Priors.pdfThomas L. Griffiths, Brian Christian, Michael L. Kalish — Proceedings of the 28th Annual Conference of the Cognitive Science SocietySun, 01 Jan 2006 00:00:00 GMT