INDEX

Explanations

mentions of the word "Labor" specifically related to government or politics

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Labor."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

irs

-0.90

imaru

-0.77

DonaldTrump

-0.68

 Flavoring

-0.65

mares

-0.65

 Allaah

-0.65

irin

-0.65

kef

-0.64

KD

-0.63

 Hiroshima

-0.62

POSITIVE LOGITS

 unions

0.85

aby

0.84

 Labor

0.80

union

0.80

Union

0.78

Labor

0.73

 union

0.71

 stopp

0.68

 ANGEL

0.67

wings

0.67

Activations Density 0.012%

mentions of the word "Labor" specifically related to government or politics

instances of the word "Labor."

No Comments

No Known Activations