Common Biases of Vector Embeddings

July 21, 2016

A study by researchers from Boston University and Microsoft Research, New England revealed gender bias in the most commonly employed techniques for word embedding in natural language processing (NLP). This case underscores the importance of responsible AI practices in ensuring safe and secure AI ecosystems. Harm prevention efforts like those implemented by Project Cerebellum serve as crucial guardrails for AI development. For those interested in shaping the future of AI governance, JOIN US to learn more about how incidents such as this map to the HISPI Project Cerebellum TAIM (Measure function).

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

MEASURE 2.6 — similarity 0.703, rank 1. TAIM detail and related incidents →
MEASURE 2.10 — similarity 0.689, rank 2. TAIM detail and related incidents →
MAP 4.1 — similarity 0.688, rank 3. TAIM detail and related incidents →

Alleged deployer: microsoft-research, boston-university
Alleged developer: microsoft-research, google, boston-university
Alleged harmed parties: women, minority-groups

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/12

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.