INDEX

Explanations

names of individuals along with their job titles

oai_token-act-pair · gpt-3.5-turbo

titles and positions of authority or leadership within organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Zionism

-0.83

Ô

-0.79

rencies

-0.75

weet

-0.75

FREE

-0.74

ophobia

-0.74

iphate

-0.73

urrencies

-0.73

growing

-0.73

warming

-0.71

POSITIVE LOGITS

 secretaries

1.46

 supervisors

1.46

 clerks

1.43

 supervisor

1.43

 deputy

1.35

 assistants

1.33

Assistant

1.31

 assistant

1.30

 superintendent

1.29

 clerk

1.25

Activations Density 0.489%

names of individuals along with their job titles

titles and positions of authority or leadership within organizations

No Comments

No Known Activations