Emergent Linear Representations in World Models of Self-Supervised Sequence Models
Authors:
Top Cited Papers
Open Access
- 1 January 2023
- conference paper
- Published by Association for Computational Linguistics (ACL)
- p. 16-30
Abstract
Neel Nanda, Andrew Lee, Martin Wattenberg. Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP. 2023.Keywords
All Related Versions
This publication has 18 references indexed in Scilit:
- Acquisition of chess knowledge in AlphaZeroProceedings of the National Academy of Sciences, 2022
- Probing Classifiers: Promises, Shortcomings, and AdvancesComputational Linguistics, 2022
- Extracting Latent Steering Vectors from Pretrained Language Models2022
- Transformer Feed-Forward Layers Are Key-Value Memories2021
- Implicit Representations of Meaning in Neural Language Models2021
- What if This Modified That? Syntactic Interventions with Counterfactual Embeddings2021
- Understanding the role of individual units in a deep neural networkProceedings of the National Academy of Sciences, 2020
- Pareto Probing: Trading Off Accuracy for Complexity2020
- Information-Theoretic Probing for Linguistic Structure2020
- A general reinforcement learning algorithm that masters chess, shogi, and Go through self-playScience, 2018