INDEX

Explanations

phrases or words related to controversial remarks or statements

oai_token-act-pair · gpt-3.5-turbo

comments and discussions about political figures and their controversial statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

neau

-0.80

inct

-0.78

fen

-0.71

Sensor

-0.71

ept

-0.69

lite

-0.69

bda

-0.68

eworks

-0.68

 CLSID

-0.68

aic

-0.67

POSITIVE LOGITS

 Mexicans

0.99

 himself

0.99

 homosexuals

0.93

 gays

0.91

 insulting

0.89

 homosexuality

0.87

 dispar

0.83

 Adolf

0.81

 apologizing

0.79

 Muslims

0.78

Activations Density 0.417%

phrases or words related to controversial remarks or statements

comments and discussions about political figures and their controversial statements

No Comments

No Known Activations