Abram Demski

Artificial Intelligence · 2016–2019

A researcher in the Science corpus: 2 papers; trusts 13 researchers and is trusted by 0 researchers here, as inferred from the citation graph.

Subfields

• Artificial Intelligence — 14 citations into it
• Safety, Robustness & Interpretability — 6 citations into it
• Probabilistic & Bayesian Methods — 5 citations into it
• Reinforcement Learning — 5 citations into it

Whom Abram Demski trusts overall

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

1. Nate Soares
2. Jessica Taylor
3. Stuart J. Russell
4. Christian P. Robert
5. S. Omohundro
6. Dan Dewey
7. C. Goodhart
8. Luc De Raedt
9. Stuart Armstrong
10. Laurent Orseau

Papers from Abram Demski's perspective

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

1. Superintelligence: Paths, Dangers, Strategies
2. The Basic AI Drives
3. GENERATIVE ADVERSARIAL NETS
4. Cooperative Inverse Reinforcement Learning
5. Concrete Problems in AI Safety
6. Corrigibility
7. Learning What to Value
8. Markov logic networks
9. Space-Time Embedded Intelligence
10. Quantilizers: A Safer Alternative to Maximizers for Limited Optimization

Their papers

• Embedded Agency (2019) — 38 citations
• Uniform Coherence (2016)