Papers by Francisco Vargas
Adversarial Concept Erasure in Kernel Space (2022.emnlp-main)
Copied to clipboard
| Challenge: | Large neural networks in NLP produce real-valued representations that encode the bit of human language that they were trained on. |
| Approach: | They propose a kernelization of the recently-proposed linear concept-removal objective and propose to remove linear subspaces from the representation space. |
| Outcome: | The proposed kernelization protects against the ability of nonlinear adversaries to recover the concept. |
Multilingual Factor Analysis (P19-1)
Copied to clipboard
| Challenge: | Existing methods for multilingual word embeddings are based on the observation that word embeds exhibit similar structures across languages. |
| Approach: | They propose a latent variable-based model that fits a multilingual dictionary to learn multilingual word representations offline. |
| Outcome: | The proposed model is robust to noise in the embedding space making it suitable for distributed representations learned from noisy corpora. |
Exploring the Linear Subspace Hypothesis in Gender Bias Mitigation (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for gender bias mitigation for word embeddings are based on pre-trained word embeds . however, the assumption that the bias subspace is linear is untested . |
| Approach: | They propose a method to isolate gender bias in word embeddings using pre-trained word embeds. |
| Outcome: | The proposed method eliminates gender bias in word embeddings but assumes bias subspace is linear . the proposed method has some drawbacks, but it is a good one for a non-linear analysis. |