LESSWRONG
LW

Wikitags

MATS Program

Edited by Ryan Kidd, Multicore, et al. last updated 30th Dec 2024

ML Alignment & Theory Scholars (MATS) Program is an educational seminar and independent research program that aims to provide talented scholars with talks, workshops, and research mentorship in the field of AI alignment, and connect them with the Berkeley AI safety research community.

Subscribe
2
Subscribe
2
Discussion0
Discussion0
Posts tagged MATS Program
682SolidGoldMagikarp (plus, prompt generation)
Ω
Jessica Rumbelow, mwatkins
3y
Ω
206
72SERI MATS Program - Winter 2022 Cohort
Ω
Ryan Kidd, Victor Warlop, Christian Smith
3y
Ω
12
334Understanding and controlling a maze-solving policy network
Ω
TurnTrout, peligrietzer, Ulisse Mini, Monte M, David Udell
2y
Ω
28
21Project proposal: Testing the IBP definition of agent
Jeremy Gillen, Thomas Larsen, JamesH
3y
4
119Soft optimization makes the value target bigger
Ω
Jeremy Gillen
3y
Ω
20
71SERI MATS - Summer 2023 Cohort
Aris, Ryan Kidd, Christian Smith
2y
25
67SERI ML Alignment Theory Scholars Program 2022
Ω
Ryan Kidd, Victor Warlop, ozhang
3y
Ω
6
63How MATS addresses “mass movement building” concerns
Ryan Kidd
2y
9
183Finite Factored Sets in Pictures
Ω
Magdalena Wache
3y
Ω
35
120Taking the parameters which seem to matter and rotating them until they don't
Garrett Baker
3y
48
26Talk: AI safety fieldbuilding at MATS
Ryan Kidd
1y
2
118Efficient Dictionary Learning with Switch Sparse Autoencoders
Anish Mudide
1y
20
105Predictions for shard theory mechanistic interpretability results
Ω
TurnTrout, Ulisse Mini, peligrietzer
3y
Ω
10
104I found >800 orthogonal "write code" steering vectors
Jacob G-W, TurnTrout
1y
19
76Neural Tangent Kernel Distillation
Thomas Larsen, Jeremy Gillen
3y
20
Load More (15/244)
Add Posts