INDEX

Explanations

phrases related to expressing opinions or giving advice

oai_token-act-pair · gpt-3.5-turbo

commas in the document

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

xxxx

-0.70

chnology

-0.67

ãģł

-0.65

ãĥīãĥ©

-0.65

ãĤº

-0.64

asaki

-0.59

circle

-0.59

ļéĨĴ

-0.59

iously

-0.56

abama

-0.56

POSITIVE LOGITS

he

1.46

she

1.28

she

0.99

 said

0.97

 according

0.89

he

0.82

said

0.82

 says

0.81

He

0.80

He

0.77

Activations Density 0.280%

phrases related to expressing opinions or giving advice

commas in the document

No Comments

No Known Activations