INDEX

Explanations

references to different social classes and their experiences or challenges

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_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.3.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.03

1:0.02

2:0.04

3:0.06

4:0.03

5:0.06

6:0.13

7:0.30

8:0.05

9:0.02

10:0.10

11:0.10

Negative Logits

��

-1.59

wagen

-1.37

 omission

-1.34

IFF

-1.28

ttp

-1.28

croft

-1.27

 captcha

-1.26

atorium

-1.26

uments

-1.26

�

-1.23

POSITIVE LOGITS

 alike

1.30

occupied

1.28

dominated

1.25

 dominated

1.22

 tribal

1.19

 lived

1.16

 retirees

1.16

 bourgeoisie

1.14

 rural

1.13

 republican

1.12

Activations Density 0.012%

references to different social classes and their experiences or challenges

No Known Activations