INDEX

Explanations

words related to legal and authoritative language

oai_token-act-pair · gpt-3.5-turbo

verbs and phrases related to declarations, statements, and understanding in legal or formal contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imeters

-0.60

isec

-0.59

spin

-0.58

 Fiesta

-0.58

pmwiki

-0.57

iami

-0.54

Companies

-0.52

taboola

-0.52

 Craigslist

-0.51

 tipped

-0.51

POSITIVE LOGITS

 that

1.20

 emphatically

1.03

 unequivocally

1.01

 furthermore

1.00

that

1.00

 moreover

0.97

:]

0.96

,—

0.90

 THAT

0.89

 firmly

0.88

Activations Density 0.341%

words related to legal and authoritative language

verbs and phrases related to declarations, statements, and understanding in legal or formal contexts

No Comments

No Known Activations