Tomasz Korbak

Artificial Intelligence · with Dymetman · 2019–2023

A researcher in the Science corpus: 9 papers; trusts 250 researchers and is trusted by 4 researchers here, as inferred from the citation graph.

Subfields

• Artificial Intelligence — 416 citations into it; #20,125 of 35,345 by trust here, #10,942 by citations
• Natural Language Processing — 210 citations into it; #3,932 of 5,688 by trust here, #4,879 by citations
• Reinforcement Learning — 128 citations into it; #1,981 of 2,242 by trust here, #2,063 by citations
• Large Language Models — 99 citations into it; #710 of 799 by trust here, #663 by citations
• Computer Vision and Pattern Recognition — 90 citations into it
• Safety, Robustness & Interpretability — 74 citations into it
• Alignment & RLHF — 63 citations into it; #52 of 62 by trust here, #29 by citations
• Evaluation & Benchmarks — 50 citations into it

Whom Tomasz Korbak trusts overall

Their strongest outbound trust, strongest first, inferred from their citations.

1. Ronald J. Williams
2. Geoffrey E. Hinton
3. Yoshua Bengio
4. Diederik P. Kingma
5. Richard S. Sutton
6. Yann LeCun
7. Ilya Sutskever
8. Radford M. Neal
9. Alex Krizhevsky
10. Alec Radford

Who trusts Tomasz Korbak overall

1. Hady ElSahar
2. Germán Kruszewski
3. Dongyoung Go
4. Marc Dymetman
5. Rafael Rafailov
6. J'er'emy Scheurer
7. Guan Wang
8. Stephen Casper
9. Kawin Ethayarajh
10. Angelica Chen

Papers from Tomasz Korbak's perspective

Ordered by how strongly the researchers Tomasz Korbak trusts endorse each paper. Same corpus, a different order for every researcher.

1. Deep Residual Learning for Image Recognition (Computer Vision and Pattern Recognition)
2. ImageNet classification with deep convolutional neural networks
3. Attention is All you Need (Neural Information Processing Systems)
4. Adam: A Method for Stochastic Optimization (International Conference on Learning Representations)
5. Long Short-Term Memory
6. Gradient-based learning applied to document recognition
7. ImageNet: A large-scale hierarchical image database
8. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (North American Chapter of the Association for Computational Linguistics)
9. Sequence to Sequence Learning with Neural Networks (Neural Information Processing Systems)
10. Neural Machine Translation by Jointly Learning to Align and Translate (International Conference on Learning Representations)

Their papers

• Pretraining Language Models with Human Preferences (2023) — 302 citations
• On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting (2022) — 94 citations
• Aligning Language Models with Preferences through f-divergence Minimization (2023) — 126 citations
• Training Language Models with Language Feedback at Scale (2023) — 138 citations
• Controlling Conditional Language Models without Catastrophic Forgetting (2021) — 43 citations
• Energy-Based Models for Code Generation under Compilability Constraints (2021) — 14 citations
• Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback (2023) — 934 citations
• RL with KL penalties is better viewed as Bayesian inference (2022) — 138 citations
• Exploiting Unsupervised Pre-training and Automated Feature Engineering for Low-resource Hate Speech Detection in Polish (2019) — 4 citations