cv
Basics
| Name | Elinor Poole-Dayan |
| elinorpd@mit.edu | |
| Url | https://elinorp-d.github.io/ |
Work
-
09/2023 - present -
05/2022 - 08/2022 Data Science Intern
Unity Technologies
- Optimized deep learning algorithms to throttle bid requests in Unity's Ad Exchange using Tensorflow.
- Decreased model training time by 25% and reduced model size and number of parameters by 50%.
- Created a text data preprocessing pipeline on Google Cloud Platform Dataflow using Apache beam.
Education
-
09/2023 - present Cambridge, MA
-
09/2019 - 05/2023 Montreal, Canada
Publications
-
11/2024 On the Relationship between Truth and Political Bias in Language Models
Empirical Methods in Natural Language Processing (EMNLP)
-
10/2024 LLM Targeted Underperformance Disproportionately Impacts Vulnerable Users
Safe Generative AI Workshop @ NeurIPS 2024
-
05/2023 Are Diffusion Models Vision-And-Language Reasoners?
Advances in Neural Information Processing Systems (NeurIPS)
-
05/2022 An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language Models
Association for Computational Linguistics
Skills
| Programming Languages | |
| Python | |
| Java | |
| SQL |
| Machine Learning | |
| Tensorflow | |
| PyTorch | |
| Keras | |
| scikit-learn | |
| pandas | |
| NumPy |
| Tools | |
| Git | |
| Docker | |
| Google Cloud Platform | |
| Amazon Web Services |
Interests
| Natural Language Processing | |
| Large Language Models | |
| Fairness & Bias | |
| Ethics in AI | |
| Alignment | |
| Human-AI Collaboration | |
| Human-Centered AI |