INDEX

Explanations

phrases expressing holding law enforcement accountable

oai_token-act-pair · gpt-3.5-turbo

repeated symbols that disrupt the text

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

 Mous

-0.74

 dispers

-0.73

 Danish

-0.68

 bearer

-0.66

 orally

-0.66

 Mayo

-0.65

 Franch

-0.65

 simultane

-0.64

 Cameroon

-0.64

 Dill

-0.64

POSITIVE LOGITS

į

1.50

ª

1.44

¤

1.43

¹

1.40

Ń

1.38

¡

1.37

Ķ

1.34

»

1.32

º

1.30

Į

1.29

Activations Density 0.105%

phrases expressing holding law enforcement accountable

repeated symbols that disrupt the text

No Comments

No Known Activations