Sotabase

Career

· PhD in Computer Science, Stanford University2023–
· Researcher, Transluce2023–
· AI/Machine Learning Intern, Apple2022–
· Research Intern, ETH Zürich2020–
· Undergraduate Student, Georgetown University2020–2024
· Admin, Wiktionary2017–
· Student Volunteer, DC Public Library
· Intern, Redwood Research

Publications (66)

ReFT: Representation Finetuning for Language Models
Neural Information Processing Systems · 2024
131
cited
Localizing Model Behavior with Path Patching
arXiv.org · 2023
125
cited
AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders
International Conference on Machine Learning · 2025
114
cited
Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability
2023
114
cited
The SIGMORPHON 2022 Shared Task on Morpheme Segmentation
Special Interest Group on Computational Morphology and Phonology Workshop · 2022
51
cited
UniMorph 4.0: Universal Morphology
International Conference on Language Resources and Evaluation · 2022
47
cited
Towards Vision-Language Mechanistic Interpretability: A Causal Tracing Tool for BLIP
2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) · 2023
46
cited
pyvene: A Library for Understanding and Improving PyTorch Models via Interventions
North American Chapter of the Association for Computational Linguistics · 2024
44
cited
CausalGym: Benchmarking causal interpretability methods on linguistic tasks
Annual Meeting of the Association for Computational Linguistics · 2024
36
cited
SIGMORPHON–UniMorph 2022 Shared Task 0: Generalization and Typologically Diverse Morphological Inflection
Special Interest Group on Computational Morphology and Phonology Workshop · 2022
36
cited
SIGMORPHON–UniMorph 2023 Shared Task 0: Typologically Diverse Morphological Inflection
Special Interest Group on Computational Morphology and Phonology Workshop · 2023
29
cited
Estimating the Entropy of Linguistic Distributions
Annual Meeting of the Association for Computational Linguistics · 2022
16
cited
Predicting positive transfer for improved low-resource speech recognition using acoustic pseudo-tokens
SIGTYP · 2024
15
cited
Bayesian scaling laws for in-context learning
arXiv.org · 2024
12
cited
Computational Historical Linguistics and Language Diversity in South Asia
Annual Meeting of the Association for Computational Linguistics · 2022
12
cited
A Reply to Makelov et al. (2023)'s "Interpretability Illusion" Arguments
arXiv.org · 2024
9
cited
PASTRIE: A Corpus of Prepositions Annotated with Supersense Tags in Reddit International English
Law · 2021
8
cited
Supervised Grapheme-to-Phoneme Conversion of Orthographic Schwas in Hindi and Punjabi
Annual Meeting of the Association for Computational Linguistics · 2020
8
cited
SNACS Annotation of Case Markers and Adpositions in Hindi
SCIL · 2021
6
cited
Improved Representation Steering for Language Models
arXiv.org · 2025
5
cited
Sotabase
Aryaman Arora | Researcher Profile | Sotabase | Sotabase