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 mini
2
3Auditing6.71.819549647668845778710
3
7Improving Feedback6.51.916884498668.573378688
4
1Force multiplier6.61.798888375855754667810
5
6Threat model evidence6.42.018.5649875963.594464779
6
2Better prediction6.32.11868635486685248899
7
4Auditing for deception6.11.64954875665794457568
8
8Informed oversight5.92.04874748445874338689
9
12Cultural shift 15.61.89464577677682354388
10
9In the loss function5.41.785745463667.554386278
11
10Norm setting5.32.19564548673782266178
12
17Forecasting discontinuities5.51.89758667456833326587
13
14Epistemic learned helplessness5.51.657.548463385644575677
14
5Enabling coordination5.41.98634444963685365589
15
16Get AIs to do it5.21.99678573435452546389
16
18Intervening on training5.12.387.584668657443312279
17
11Regulation5.12.226.5343342568744454910
18
13Cultural shift 24.91.73464358367553345387
19
15Microscope AI5.02.003743392456.53585476
20
20ELK5.32.3068337543237986
21
19Auditing a training run4.92.225.584468254732145488
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