INDEX

Explanations

mentions of specific individuals involved in extremist activities

oai_token-act-pair · gpt-3.5-turbo

names of individuals and groups associated with conflict or violence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Dragonbound

-0.79

Voy

-0.79

 Uncharted

-0.79

 Deadpool

-0.78

 GDDR

-0.76

 Titanic

-0.73

 Wilde

-0.73

 mammal

-0.73

FTC

-0.72

 Victorian

-0.71

POSITIVE LOGITS

awi

1.10

qqa

1.06

Islamic

1.06

iyah

1.06

abi

1.03

 Islam

1.03

 mosque

1.00

 Islamist

0.99

 Ahmad

0.98

aida

0.97

Activations Density 0.287%

mentions of specific individuals involved in extremist activities

names of individuals and groups associated with conflict or violence

No Comments

No Known Activations