Go Back
Choose this branch
Choose this branch
meritocratic
regular
democratic
hot
top
alive
1564 posts
AI
Inner Alignment
Interpretability (ML & AI)
AI Timelines
GPT
Research Agendas
AI Takeoff
Value Learning
Machine Learning (ML)
Conjecture (org)
Mesa-Optimization
Outer Alignment
349 posts
Abstraction
Impact Regularization
Rationality
World Modeling
Decision Theory
Human Values
Goal-Directedness
Anthropics
Utility Functions
Finite Factored Sets
Shard Theory
Fixed Point Theorems
28
Discovering Language Model Behaviors with Model-Written Evaluations
evhub
4h
3
13
Note on algorithms with multiple trained components
Steven Byrnes
7h
1
45
Towards Hodge-podge Alignment
Cleo Nardo
1d
20
30
Take 12: RLHF's use is evidence that orgs will jam RL at real-world problems.
Charlie Steiner
19h
0
35
The "Minimal Latents" Approach to Natural Abstractions
johnswentworth
22h
6
11
An Open Agency Architecture for Safe Transformative AI
davidad
11h
11
213
AI alignment is distinct from its near-term applications
paulfchristiano
7d
5
114
How "Discovering Latent Knowledge in Language Models Without Supervision" Fits Into a Broader Alignment Scheme
Collin
5d
18
73
Can we efficiently explain model behaviors?
paulfchristiano
4d
0
64
Paper: Constitutional AI: Harmlessness from AI Feedback (Anthropic)
LawrenceC
4d
10
99
Trying to disambiguate different questions about whether RLHF is “good”
Buck
6d
39
23
Event [Berkeley]: Alignment Collaborator Speed-Meeting
AlexMennen
1d
2
27
Take 11: "Aligning language models" should be weirder.
Charlie Steiner
2d
0
55
High-level hopes for AI alignment
HoldenKarnofsky
5d
3
78
Shard Theory in Nine Theses: a Distillation and Critical Appraisal
LawrenceC
1d
9
46
Positive values seem more robust and lasting than prohibitions
TurnTrout
3d
9
125
Finite Factored Sets in Pictures
Magdalena Wache
9d
29
59
Take 7: You should talk about "the human's utility function" less.
Charlie Steiner
12d
22
58
Alignment allows "nonrobust" decision-influences and doesn't require robust grading
TurnTrout
21d
27
573
Where I agree and disagree with Eliezer
paulfchristiano
6mo
205
48
Open technical problem: A Quinean proof of Löb's theorem, for an easier cartoon guide
Andrew_Critch
26d
34
75
«Boundaries», Part 3a: Defining boundaries as directed Markov blankets
Andrew_Critch
1mo
13
67
Humans do acausal coordination all the time
Adam Jermyn
1mo
36
95
Contra shard theory, in the context of the diamond maximizer problem
So8res
2mo
16
239
Without specific countermeasures, the easiest path to transformative AI likely leads to AI takeover
Ajeya Cotra
5mo
89
155
The shard theory of human values
Quintin Pope
3mo
57
64
Notes on "Can you control the past"
So8res
2mo
40
40
Counterfactability
Scott Garrabrant
1mo
4