INDEX

Explanations

words related to fire incidents or activities involving fire

oai_token-act-pair · gpt-3.5-turbo

references to fire incidents or related occurrences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Vide

-0.79

sembly

-0.75

berra

-0.75

Tok

-0.72

atem

-0.71

xus

-0.69

ĸļ

-0.68

 fixme

-0.67

 Citizen

-0.66

 Freed

-0.66

POSITIVE LOGITS

flies

1.17

storm

1.01

proof

1.01

 exting

0.98

storms

0.92

 brigade

0.87

bomb

0.87

balls

0.85

fly

0.85

 retard

0.84

Activations Density 0.021%

words related to fire incidents or activities involving fire

references to fire incidents or related occurrences

No Comments

No Known Activations