lukemarks

Karma: 466

Steering Language Models in Multiple Directions Simultaneously

lukemarks, Narmeen and Amirali Abdullah

May 2, 2025, 3:27 PM

18 points

0 comments7 min readLW link

lukemarks’s Shortform

lukemarksJul 2, 2024, 6:56 AM

4 points

19 comments LW link

Beta Tester Request: Rallypoint Bounties

lukemarksMay 25, 2024, 9:11 AM

25 points

4 comments1 min readLW link

[Question] Shouldn’t we ‘Just’ Superimitate Low-Res Uploads?

lukemarksNov 3, 2023, 7:42 AM

15 points

2 comments2 min readLW link

Early Experiments in Reward Model Interpretation Using Sparse Autoencoders

lukemarks, Amirali Abdullah, Rauno Arike, Fazl and nothoughtsheadempty

Oct 3, 2023, 7:45 AM

17 points

0 comments5 min readLW link

The Löbian Obstacle, And Why You Should Care

lukemarksSep 7, 2023, 11:59 PM

18 points

6 comments2 min readLW link

[Question] What Does LessWrong/EA Think of Human Intelligence Augmentation as of mid-2023?

lukemarksJul 8, 2023, 11:42 AM

84 points

28 comments2 min readLW link

Direct Preference Optimization in One Minute

lukemarksJun 26, 2023, 11:52 AM

22 points

3 comments2 min readLW link

Partial Simulation Extrapolation: A Proposal for Building Safer Simulators

lukemarksJun 17, 2023, 1:55 PM

16 points

0 comments10 min readLW link

Higher Dimension Cartesian Objects and Aligning ‘Tiling Simulators’

lukemarksJun 11, 2023, 12:13 AM

22 points

0 comments5 min readLW link

The Security Mindset, S-Risk and Publishing Prosaic Alignment Research

lukemarksApr 22, 2023, 2:36 PM

40 points

7 comments5 min readLW link

Select Agent Specifications as Natural Abstractions

lukemarksApr 7, 2023, 11:16 PM

19 points

3 comments5 min readLW link