ABCDEFGHIJKLMNOPQRSTUVWXYZAAABACAD
1
s NameMeanStdNeel NandaEvan HubingerAdam ScherlisVictoria KrakovnaJanos KramarDane SherburnCharlie SteinerNick TurnerRamana KumarSimeon Campos
Aron Malmborg
Jenny Nitishinskaya
Charbel-Raphael Segerie
Arthur ConmyCurt TiggesMark BissellClaude HaikuGPT-4o miniSebastian Jost
2
3Auditing6.71.7795496476688457787107.5
3
7Improving Feedback6.51.866884498668.5733786886
4
1Force multiplier6.61.7788883758557546678108
5
6Threat model evidence6.62.048.5649875963.5944647799
6
2Better prediction6.32.068686354866852488997
7
4Auditing for deception6.21.619548756657944575687
8
8Informed oversight5.91.998747484458743386895.5
9
12Cultural shift 15.51.844645776776823543885
10
9In the loss function5.31.735745463667.5543862785
11
10Norm setting5.32.135645486737822661785.5
12
17Forecasting discontinuities5.51.847586674568333265876
13
14Epistemic learned helplessness5.41.647.5484633856445756774
14
5Enabling coordination5.32.006344449636853655893
15
16Get AIs to do it5.12.006785734354525463893
16
18Intervening on training5.32.407.5846686574433122798
17
11Regulation5.12.166.53433425687444549105
18
13Cultural shift 24.91.704643583675533453874
19
15Microscope AI4.92.003743392456.535854763
20
20ELK5.42.26683375432379867
21
19Auditing a training run5.12.275.5844682547321454888
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100