INDEX

Explanations

words related to fires and emergencies

oai_token-act-pair · gpt-3.5-turbo

mentions of fires or fire-related incidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

xus

-0.83

laus

-0.74

 Vide

-0.74

 Freed

-0.73

 Lans

-0.72

atem

-0.72

 Virtue

-0.72

sembly

-0.70

amaru

-0.68

VIDIA

-0.68

POSITIVE LOGITS

flies

1.16

storm

1.14

 exting

1.10

proof

1.07

balls

1.02

storms

1.02

fighting

1.00

fly

1.00

trap

0.98

fight

0.96

Activations Density 0.029%

words related to fires and emergencies

mentions of fires or fire-related incidents

No Comments

No Known Activations