AI ALIGNMENT FORUMTags
AF

Refine

Help improve this page

Help improve this page

Contributors

Refine is a conceptual research incubator hosted by Conjecture.

Posts tagged Refine

4

16confusion about alignment requirements

8mo

0

1

11Refine's Second Blog Post Day

9mo

0

1

16What if we approach AI safety like a technical engineering safety problem

9mo

0

1

17Benchmarking Proposals on Risk Scenarios

9mo

0

2

48What I Learned Running Refine

6mo

3

1

24PreDCA: vanessa kosoy's alignment protocol

9mo

1

1

21the Insulated Goal-Program idea

10mo

3

2

17Representational Tethers: Tying AI Latents To Human Ones

8mo

0

0

21Epistemic Artefacts of (conceptual) AI alignment research

Nora_Ammann, particlemania

9mo

0

2

16ordering capability thresholds

8mo

0

1

14goal-program bricks

10mo

0

1

13Refine Blogpost Day #3: The shortforms I did write

Alexander Gietelink Oldenziel

8mo

0

2

9Refine's Third Blog Post Day/Week

8mo

0

0

44Refine: An Incubator for Conceptual Alignment Research Bets

1y

4

0

29How to Diversify Conceptual Alignment: the Model Behind Refine

10mo

2