Vafa Ghazavi

Safety Research · 2021

A researcher in the Science corpus: 1 paper; trusts 8 researchers and is trusted by 0 researchers here, as inferred from the citation graph.

Subfields

• Safety Research — 13 citations into it
• Safety, Robustness & Interpretability — 12 citations into it
• Artificial Intelligence — 11 citations into it

Whom Vafa Ghazavi trusts overall

Their strongest outbound trust, strongest first, inferred from their citations.

1. Paul Christiano
2. Joy Buolamwini
3. Jan Leike
4. Dario Amodei
5. Timnit Gebru
6. Tom Everitt

Papers from Vafa Ghazavi's perspective

Ordered by how strongly the researchers Vafa Ghazavi trusts endorse each paper. Same corpus, a different order for every researcher.

1. Concrete Problems in AI Safety
2. AI safety via debate
3. Scalable agent alignment via reward modeling: a research direction (arXiv.org)
4. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification
5. Supervising strong learners by amplifying weak experts
6. ImageNet classification with deep convolutional neural networks
7. To predict and serve?
8. Artificial Intelligence, Values, and Alignment
9. Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor
10. Adam: A Method for Stochastic Optimization (International Conference on Learning Representations)

Their papers

• The Challenge of Value Alignment: from Fairer Algorithms to AI Safety (2021) — 71 citations