x
This website requires javascript to properly function. Consider activating javascript to get access to all site functionality.
LessWrong & Lighthaven are fundraising for 2026
– our goal is to raise $2M for the next year
Donate
LESSWRONG
LW
Login
Refine — LessWrong
Refine
Edited by
Ruby
,
Mateusz Bagiński
last updated
16th Nov 2025
Refine
was
a conceptual research incubator hosted by
Conjecture
.
Subscribe
Discussion
Subscribe
Discussion
Posts tagged
Refine
Most Relevant
4
19
Refine's Second Blog Post Day
Ω
adamShimi
3y
Ω
0
3
36
What if we approach AI safety like a technical engineering safety problem
Ω
zeshen
3y
Ω
4
3
25
Benchmarking Proposals on Risk Scenarios
Ω
Paul Bricman
3y
Ω
2
2
108
What I Learned Running Refine
Ω
adamShimi
3y
Ω
5
2
31
Epistemic Artefacts of (conceptual) AI alignment research
Ω
Nora_Ammann
,
particlemania
3y
Ω
1
2
30
Representational Tethers: Tying AI Latents To Human Ones
Ω
Paul Bricman
3y
Ω
0
2
23
Refine Blogpost Day #3: The shortforms I did write
Ω
Alexander Gietelink Oldenziel
3y
Ω
0
2
18
Refine's Third Blog Post Day/Week
Ω
adamShimi
3y
Ω
0
1
144
Refine: An Incubator for Conceptual Alignment Research Bets
Ω
adamShimi
4y
Ω
13
1
87
How to Diversify Conceptual Alignment: the Model Behind Refine
Ω
adamShimi
3y
Ω
11
1
55
Refine's First Blog Post Day
Ω
adamShimi
3y
Ω
3
1
54
All the posts I will never write
Ω
Alexander Gietelink Oldenziel
3y
Ω
8
1
53
I missed the crux of the alignment problem the whole time
Ω
zeshen
3y
Ω
7
1
50
My Thoughts on the ML Safety Course
Ω
zeshen
3y
Ω
3
1
42
A newcomer’s guide to the technical AI safety field
Ω
zeshen
3y
Ω
3