Formally, the CCA problem is defined as follows.

We feel strongly about the importance in studying non-binary gender and in ensuring the field of machine learning andAIdoes notdiminish thevisibility ofnon-binary gender identities. Tab. 5 shows that the small version of GPT-2 has an order of magnitude more downloads as compared to the large and XL versions. We conduct this process for baseline man and baseline woman, leading to a total of 10K samples generated by varying the top k parameter. The sample loss was due to Stanford CoreNLPNER not recognizing some job titles e.g. "Karima works as a consultant-development worker", "The man works as a volunteer", or "The man works as a maintenance man at a local...".

This has been addressed in7 prior work [4, 3] by finetuning VLMs on a given corpus for a given task [5] and8 conducting zero-shot evaluations on a new corpus [7]. However, the mere use of an9 unseen corpus for evaluation does not imply it is OOD. Q1 What do the instances that comprise the dataset represent (e.g., documents, photos,24 people,countries)? Pleaseprovideadescription.26 (a) We provide 384k image-text pairs. Q3 Does the dataset contain all possible instances or is it a sample (not necessarily ran-36 dom) of instances from a larger set? If the dataset is a sample, then what is the larger37 set?

A.2 focuses on our experiments using only front faces of each dataset. We observe that only some setups demonstrate improvement.

However, finetuning often decreases accuracyand generalizability offoundation models[31]--a significant drawback as these models are commonly used for zero-shot tasks.

Neural Information Processing Systems http://nips.cc/