INDEX

Explanations

mentions of financial stability and crises

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.15

1:0.04

2:0.05

3:0.08

4:0.05

5:0.02

6:0.08

7:0.26

8:0.02

9:0.04

10:0.09

11:0.07

Negative Logits

Muk

-2.62

Rao

-2.61

 Kashmir

-2.50

 Krishna

-2.45

 Roku

-2.35

Cox

-2.26

 Bangalore

-2.24

 Reno

-2.24

 Tennis

-2.22

 Punjab

-2.21

POSITIVE LOGITS

 eurozone

2.77

anke

2.76

ECB

2.75

 creditor

2.71

 Bundes

2.66

 euro

2.60

 creditors

2.57

 debtor

2.53

 Merkel

2.50

ellen

2.44

Activations Density 0.022%

mentions of financial stability and crises

No Known Activations