INDEX

Explanations

keywords related to explosive devices

oai_token-act-pair · gpt-3.5-turbo

mentions of explosives or bombing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

laus

-0.83

éĸ

-0.76

SEA

-0.76

TH

-0.74

é»Ĵ

-0.74

DAY

-0.71

VID

-0.71

tyard

-0.70

CE

-0.70

Lic

-0.69

POSITIVE LOGITS

arding

1.16

arded

1.10

 bomb

1.03

astically

0.98

 detonated

0.97

 deton

0.96

astic

0.93

 bombs

0.92

 blasts

0.90

 shelters

0.87

Activations Density 0.010%

keywords related to explosive devices

mentions of explosives or bombing

No Comments

No Known Activations