INDEX

Explanations

names or references associated with notable individuals or titles in political contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

0:0.02

1:0.02

2:0.04

3:0.04

4:0.03

5:0.03

6:0.43

7:0.15

8:0.04

9:0.06

10:0.05

11:0.03

Negative Logits

comed

-1.43

APP

-1.31

Newsletter

-1.26

igmat

-1.25

abytes

-1.22

ITNESS

-1.21

sonian

-1.19

 Latvia

-1.18

 exceed

-1.17

 Estonia

-1.16

POSITIVE LOGITS

again

1.33

yden

1.32

helial

1.32

oxide

1.32

eller

1.32

 cuff

1.30

isner

1.27

ondo

1.24

neau

1.23

oy

1.22

Activations Density 0.003%

names or references associated with notable individuals or titles in political contexts

No Comments

No Known Activations