INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.04

1:0.04

2:0.09

3:0.10

4:0.22

5:0.04

6:0.09

7:0.05

8:0.05

9:0.07

10:0.08

11:0.07

Negative Logits

�

-1.66

EXT

-1.50

ench

-1.49

Appearance

-1.48

 Bung

-1.38

�

-1.36

-1.34

sett

-1.34

-1.33

deg

-1.30

POSITIVE LOGITS

 monarch

1.49

 sovereign

1.37

essor

1.34

earch

1.34

ikhail

1.33

 Lumpur

1.31

cum

1.30

 coun

1.28

EGA

1.27

 millenn

1.27

Activations Density 0.000%

No Known Activations

This feature has no known activations.