AI ALIGNMENT FORUM
Wikitags
AF

Academic Papers

Written by Kaj Sotala last updated 9th Jul 2020

Posts either linking to, or summarizing, formal papers published elsewhere.

Posts tagged Academic Papers

1

64Some AI research areas and their relevance to existential safety

4y

34

2

19How to Control an LLM's Behavior (why my P(DOOM) went down)

Roger Dearnaley

1y

0

2

512021 AI Alignment Literature Review and Charity Comparison

3y

15

1

59Evidence of Learned Look-Ahead in a Chess-Playing Neural Network

10mo

0

1

29A technical note on bilinear layers for interpretability

2y

0

1

19Formal Solution to the Inner Alignment Problem

4y

123

1

21Why is pseudo-alignment "worse" than other ways ML can fail to generalize?

nostalgebraist, Evan Hubinger

5y

7

0

69Shallow review of technical AI safety, 2024

technicalities, Stag, Stephen McAleese, jordine, Dr. David Mathers

3mo

1

1

28NeurIPS ML Safety Workshop 2022

3y

1

1

25How truthful is GPT-3? A benchmark for language models

4y

18

0

17Secret Collusion: Will We Know When to Unplug AI?

schroederdewitt, srm, Mikhail Baranchuk, Lewis Hammond, chansmi, Angira Sharma

6mo

1

0

14New paper: Corrigibility with Utility Preservation

6y

9

0

18Learning preferences by looking at the world

6y

4

0

26Human-AI Collaboration

5y

4

0

19Learning biases and rewards simultaneously

6y

3