INDEX

Explanations

monetary values and pricing information

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

Head Attr Weights

0:0.02

1:0.02

2:0.11

3:0.13

4:0.04

5:0.03

6:0.05

7:0.12

8:0.04

9:0.06

10:0.23

11:0.09

Negative Logits

cknow

-1.78

cknowled

-1.60

accompanied

-1.50

lia

-1.47

upon

-1.44

called

-1.44

 repeated

-1.41

ablish

-1.39

leaders

-1.39

 surrog

-1.36

POSITIVE LOGITS

wcsstore

1.62

 stock

1.60

sale

1.59

 premium

1.53

bucks

1.47

license

1.46

 pric

1.45

stocks

1.45

Price

1.43

Stock

1.42

Activations Density 0.001%

monetary values and pricing information

No Known Activations