János Kramár

33 papers A* 6A 2Journal 20Unranked 5

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	Building Production-Ready Probes For Gemini. CoRR János Kramár, Joshua Engels, Zheng Wang, Bilal Chughtai, Rohin Shah, Neel Nanda, Arthur Conmy
2024	J	`jnl`	AtP*: An efficient and scalable method for localizing LLM behaviour to components. CoRR János Kramár, Tom Lieberum, Rohin Shah, Neel Nanda
2024	J	`jnl`	Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2. CoRR Tom Lieberum, Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Nicolas Sonnerat, Vikrant Varma, János Kramár, Anca D. Dragan, Rohin Shah, Neel Nanda
2024	J	`jnl`	Improving Dictionary Learning with Gated Sparse Autoencoders. CoRR Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Tom Lieberum, Vikrant Varma, János Kramár, Rohin Shah, Neel Nanda
2024	A*	`conf`	Improving Sparse Decomposition of Language Model Activations with Gated Sparse Autoencoders. NeurIPS Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Tom Lieberum, Vikrant Varma, János Kramár, Rohin Shah, Neel Nanda
2024	J	`jnl`	Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders. CoRR Senthooran Rajamanoharan, Tom Lieberum, Nicolas Sonnerat, Arthur Conmy, Vikrant Varma, János Kramár, Neel Nanda
2024	A*	`conf`	On scalable oversight with weak LLMs judging strong LLMs. NeurIPS Zachary Kenton, Noah Y. Siegel, János Kramár, Jonah Brown-Cohen, Samuel Albanie, Jannis Bulian, Rishabh Agarwal, David Lindner, Yunhao Tang, Noah D. Goodman, Rohin Shah
2024	J	`jnl`	On scalable oversight with weak LLMs judging strong LLMs. CoRR Zachary Kenton, Noah Y. Siegel, János Kramár, Jonah Brown-Cohen, Samuel Albanie, Jannis Bulian, Rishabh Agarwal, David Lindner, Yunhao Tang, Noah D. Goodman, Rohin Shah
2023	J	`jnl`	Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla. CoRR Tom Lieberum, Matthew Rahtz, János Kramár, Neel Nanda, Geoffrey Irving, Rohin Shah, Vladimir Mikulik
2023	J	`jnl`	Explaining grokking through circuit efficiency. CoRR Vikrant Varma, Rohin Shah, Zachary Kenton, János Kramár, Ramana Kumar
2023	J	`jnl`	Power-seeking can be probable and predictive for trained agents. CoRR Victoria Krakovna, János Kramár
2023	J	`jnl`	The Hydra Effect: Emergent Self-repair in Language Model Computations. CoRR Thomas McGrath, Matthew Rahtz, János Kramár, Vladimir Mikulik, Shane Legg
2023	A*	`conf`	Tracr: Compiled Transformers as a Laboratory for Interpretability. NeurIPS David Lindner, János Kramár, Sebastian Farquhar, Matthew Rahtz, Tom McGrath, Vladimir Mikulik
2023	J	`jnl`	Tracr: Compiled Transformers as a Laboratory for Interpretability. CoRR David Lindner, János Kramár, Matthew Rahtz, Thomas McGrath, Vladimir Mikulik
2022	A	`conf`	Sample-based Approximation of Nash in Large Many-Player Games via Gradient Descent. AAMAS Ian Gemp, Rahul Savani, Marc Lanctot, Yoram Bachrach, Thomas W. Anthony, Richard Everett, Andrea Tacchetti, Tom Eccles, János Kramár
2021	A*	`conf`	A Neural Network Auction For Group Decision Making Over a Continuous Space. IJCAI Yoram Bachrach, Ian Gemp, Marta Garnelo, János Kramár, Tom Eccles, Dan Rosenbaum, Thore Graepel
2021	J	`jnl`	Sample-based Approximation of Nash in Large Many-Player Games via Gradient Descent. CoRR Ian Gemp, Rahul Savani, Marc Lanctot, Yoram Bachrach, Thomas W. Anthony, Richard Everett, Andrea Tacchetti, Tom Eccles, János Kramár
2020	A*	`conf`	Learning to Play No-Press Diplomacy with Best Response Policy Iteration. NeurIPS Thomas W. Anthony, Tom Eccles, Andrea Tacchetti, János Kramár, Ian Gemp, Thomas C. Hudson, Nicolas Porcel, Marc Lanctot, Julien Pérolat, Richard Everett, Satinder Singh, Thore Graepel, Yoram Bachrach
2020	J	`jnl`	Learning to Play No-Press Diplomacy with Best Response Policy Iteration. CoRR Thomas W. Anthony, Tom Eccles, Andrea Tacchetti, János Kramár, Ian Gemp, Thomas C. Hudson, Nicolas Porcel, Marc Lanctot, Julien Pérolat, Richard Everett, Satinder Singh, Thore Graepel, Yoram Bachrach
2020	—	`conf`	Should I Tear down This Wall? Optimizing Social Metrics by Evaluating Novel Actions. COIN(E)@AAMAS János Kramár, Neil C. Rabinowitz, Tom Eccles, Andrea Tacchetti
2020	J	`jnl`	Should I tear down this wall? Optimizing social metrics by evaluating novel actions. CoRR János Kramár, Neil C. Rabinowitz, Tom Eccles, Andrea Tacchetti
2019	J	`jnl`	Learning Reciprocity in Complex Sequential Social Dilemmas. CoRR Tom Eccles, Edward Hughes, János Kramár, Steven Wheelwright, Joel Z. Leibo
2019	J	`jnl`	OpenSpiel: A Framework for Reinforcement Learning in Games. CoRR Marc Lanctot, Edward Lockhart, Jean-Baptiste Lespiau, Vinícius Flores Zambaldi, Satyaki Upadhyay, Julien Pérolat, Sriram Srinivasan, Finbarr Timbers, Karl Tuyls, Shayegan Omidshafiei, Daniel Hennes, Dustin Morrill, Paul Muller, Timo Ewalds, Ryan Faulkner, János Kramár, Bart De Vylder, Brennan Saeta, James Bradbury, David Ding, Sebastian Borgeaud, Matthew Lai, Julian Schrittwieser, Thomas W. Anthony, Edward Hughes, Ivo Danihelka, Jonah Ryan-Davis
2019	—	`conf`	Relational Forward Models for Multi-Agent Learning. ICLR (Poster) Andrea Tacchetti, H. Francis Song, Pedro A. M. Mediano, Vinícius Flores Zambaldi, János Kramár, Neil C. Rabinowitz, Thore Graepel, Matthew M. Botvinick, Peter W. Battaglia
2019	A	`conf`	The Imitation Game: Learned Reciprocity in Markov games. AAMAS Tom Eccles, Edward Hughes, János Kramár, Steven Wheelwright, Joel Z. Leibo
2018	—	`conf`	Reinforcement and Imitation Learning for Diverse Visuomotor Skills. Robotics: Science and Systems Yuke Zhu, Ziyu Wang, Josh Merel, Andrei A. Rusu, Tom Erez, Serkan Cabi, Saran Tunyasuvunakool, János Kramár, Raia Hadsell, Nando de Freitas, Nicolas Heess
2018	J	`jnl`	Reinforcement and Imitation Learning for Diverse Visuomotor Skills. CoRR Yuke Zhu, Ziyu Wang, Josh Merel, Andrei A. Rusu, Tom Erez, Serkan Cabi, Saran Tunyasuvunakool, János Kramár, Raia Hadsell, Nando de Freitas, Nicolas Heess
2017	J	`jnl`	Guidelines for Artificial Intelligence Containment. CoRR James Babcock, János Kramár, Roman V. Yampolskiy
2017	—	`conf`	Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations. ICLR (Poster) David Krueger, Tegan Maharaj, János Kramár, Mohammad Pezeshki, Nicolas Ballas, Nan Rosemary Ke, Anirudh Goyal, Yoshua Bengio, Aaron C. Courville, Christopher J. Pal
2016	—	`conf`	The AGI Containment Problem. AGI James Babcock, János Kramár, Roman Yampolskiy
2016	J	`jnl`	The AGI Containment Problem. CoRR James Babcock, János Kramár, Roman Yampolskiy
2016	J	`jnl`	Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations. CoRR David Krueger, Tegan Maharaj, János Kramár, Mohammad Pezeshki, Nicolas Ballas, Nan Rosemary Ke, Anirudh Goyal, Yoshua Bengio, Hugo Larochelle, Aaron C. Courville, Chris Pal
2010	A*	`conf`	A Generalized-Zero-Preserving Method for Compact Encoding of Concept Lattices. ACL Matthew Skala, Victoria Krakovna, János Kramár, Gerald Penn