Articles by Tomasz Korbak

Claim page

Tomasz Korbak

Activity

Upvotes

Folders

Articles

The Two-Hop Curse: LLMs trained on A\rightarrowB, B\rightarrowC fail to learn A\rightarrowC

6 January 2025 by Mikita Balesni and others

Computation and Language, Artificial Intelligence

Safety case template for frontier AI: A cyber inability argument

12 November 2024 by Arthur Goemans and others

Computers and Society, Cryptography and Security

Towards evaluations-based safety cases for AI scheming

7 November 2024 by Mikita Balesni and others

Cryptography and Security, Artificial Intelligence

Looking Inward: Language Models Can Learn About Themselves by Introspection

17 October 2024 by Felix Binder and others

Computation and Language, Artificial Intelligence

Foundational Challenges in Assuring Alignment and Safety of Large Language Models

6 September 2024 by Usman Anwar and others at

MIT

Machine Learning, Artificial Intelligence

The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"

26 May 2024 by Lukas Berglund and others

Computation and Language, Artificial Intelligence

Inverse Scaling: When Bigger Isn't Better

13 May 2024 by Ian Mckenzie and others

Computation and Language, Artificial Intelligence

Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data

29 April 2024 by Matthias Gerstgrasser and others at

Stanford University

Machine Learning, Artificial Intelligence

Aligning language models with human preferences

18 April 2024 by Tomasz Korbak

Machine Learning, Computation and Language

Catalytic Role Of Noise And Necessity Of Inductive Biases In The Emergence Of Compositional Communication

3 April 2024 by Łukasz Kuciński and others

Machine Learning, Artificial Intelligence

Compositional preference models for aligning LMs

14 March 2024 by Dongyoung Go and others

Computation and Language, Machine Learning

Improving Code Generation by Training with Natural Language Feedback

22 February 2024 by Angelica Chen and others

Software Engineering, Artificial Intelligence

Training Language Models with Language Feedback at Scale

22 February 2024 by Jérémy Scheurer and others

Computation and Language, Artificial Intelligence

Towards Understanding Sycophancy in Language Models

27 October 2023 by Mrinank Sharma and others

Computation and Language, Artificial Intelligence

Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback

11 September 2023 by Stephen Casper and others

Artificial Intelligence, Computation and Language

Taken out of context: On measuring situational awareness in LLMs

1 September 2023 by Lukas Berglund and others

Computation and Language, Machine Learning

Pretraining Language Models with Human Preferences

14 June 2023 by Tomasz Korbak and others

Computation and Language, Machine Learning

Aligning Language Models with Preferences through f-divergence Minimization

6 June 2023 by Dongyoung Go and others

Computation and Language, Machine Learning

Models of symbol emergence in communication: a conceptual review and a guide for avoiding local minima

8 March 2023 by Julian Zubek and others

Artificial Intelligence, Computation and Language

On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting

14 November 2022 by Tomasz Korbak and others

Machine Learning, Computation and Language

RL with KL penalties is better viewed as Bayesian inference

21 October 2022 by Tomasz Korbak and others

Machine Learning

Controlling Conditional Language Models without Catastrophic Forgetting

20 June 2022 by Tomasz Korbak and others

Machine Learning, Computation and Language

A continuity of Markov blanket interpretations under the Free Energy Principle

18 January 2022 by Anil Seth and others

Neurons and Cognition

Energy-Based Models for Code Generation under Compilability Constraints

9 June 2021 by Tomasz Korbak and others

Machine Learning, Computation and Language

Measuring non-trivial compositionality in emergent communication

29 October 2020 by Tomasz Korbak and others

Neural and Evolutionary Computing, Computation and Language

Developmentally motivated emergence of compositional communication via template transfer

4 October 2019 by Tomasz Korbak and others

Machine Learning, Artificial Intelligence

Exploiting Unsupervised Pre-training and Automated Feature Engineering for Low-resource Hate Speech Detection in Polish

17 June 2019 by Renard Korzeniowski and others

Computation and Language, Machine Learning

Fine-tuning Tree-LSTM for phrase-level sentiment classification on a Polish dependency treebank. Submission to PolEval task 2

3 November 2017 by Tomasz Korbak and Paulina Żak

Computation and Language

This is an AI-generated summary

Key points

Topics

Machine Learning Computation and Language Artificial Intelligence Computers and Society Software Engineering Multiagent Systems Neural and Evolutionary Computing Emerging Technologies Neurons and Cognition