Papers
Conference and workshop papers
- Korbak, T., Shi, K., Chen, A., Bhalerao, R., Buckley, C. Phang, J., Bowman S. & Perez. E. (2023). Pretraining Language Models with Human Preferences.
- Scheurer, J., Ander Campos, J., Korbak, T., Jun Shern, C., Chen, A., Cho, K., Perez, E. (2023). Training Language Models with Language Feedback at Scale.
- Chen, A., Scheurer, J., Korbak, T., Ander Campos, J., Jun Shern, C., Bowman, S., Cho, K., Perez, E. (2023). Improving Code Generation by Training with Natural Language Feedback.
- Go, D., Korbak, T. Rozen, J., Ryu, N., Kruszewski, G. Dymetman, M. (2023). Aligning Language Models with Preferences through f-divergence Minimization.
- Korbak, T., Elsahar, H., Kruszewski, G. & Dymetman, M. (2022). On reinforcement learning and distribution matching for fine-tuning language models with no catastrophic forgetting. NeurIPS 2022.
- Korbak, T., Elsahar, H., Kruszewski, G. & Dymetman, M. (2022). Controlling conditional language models without catastrophic forgetting. ICML 2022.
- Korbak, T., Perez, E. & Buckley, C. (2022). RL with KL penalties is better viewed as Bayesian inference. Findings in EMNLP 2022.
- Kuciński, Ł., Korbak, T., Kołodziej, P. & Miłoś, P. (2021). Catalytic role of noise and necessity of inductive biases in emergence of compositional communication. NeurIPS 2021.
- Korbak, T., Elsahar, H., Dymetman, M. & Kruszewski, G. Energy-based models for code generation under compilability constraints. ACL 2021 workshop on NLP for Programming.
- Korbak, T., Zubek, J. & Rączaszek-Leonardi, J. (2020). Measuring non-trivial compositionality in emergent communication. NeurIPS 2020 workshop on Emergent Communication.
- Korbak, T., Zubek, J., Kuciński, Ł., Miłoś, P. & Rączaszek-Leonardi, J. (2019). Developmentally motivated emergence of compositional communication via template transfer. NeurIPS 2019 workshop on Emergent Communication.
- Korzeniowski, R., Rolczyński, R., Sadownik, P., Korbak, T. & Możejko, M. (2019). Exploiting Unsupervised Pre-training and Automated Feature Engineering for Low-resource Hate Speech Detection in Polish. Proceedings of the PolEval 2019 Workshop.
- Korbak, T. & Żak, P. (2017). Fine-tuning Tree-LSTM for phrase-level sentiment classification on a Polish dependency treebank. In Z. Vetulani and P. Paroubek (eds.) Proceedings of the 8th Language & Technology Conference.
Journal papers
- Rorot, W., Korbak, T., Litwin, P. & Miłkowski, M. (2022). Enough blanket metaphysics, time for data-driven heuristics. Behavioral and Brain Sciences.
- Seth, A., Korbak, T. & Tschantz, A. (2022). A continuity of Markov blanket interpretations under the Free Energy Principle. Behavioral and Brain Sciences.
- Korbak, T., Zubek, J., Kuciński, Ł., Miłoś & P. & Rączaszek-Leonardi, J. (2022). Interaction history as a source of compositionality in emergent communication. Interaction Studies.
- Korbak, T. (2022). Self-organisation, (M, R)–systems and enactive cognitive science. Adaptive Behavior.
- Korbak, T. (2019). Computational enactivism under the free energy principle. Synthese.
- Korbak, T. (2019). Unsupervised learning and the natural origins of content. Avant.
- Korbak, T. (2015). Scaffolded minds and the evolution of content in signaling pathways. Studies in Logic, Grammar and Rhetoric, 41 (54).
- Korbak, T. (2015). Apercepcja transcendentalna w kantowskim modelu epigenezy czystego rozumu. Przegląd Filozoficzny – Nowa seria, 3 (95).