INDEX

Explanations

phrases related to setting up and creating something in a technical or developmental context

oai_token-act-pair · gpt-3.5-turbo

instructions and actions related to setting up and developing software environments or applications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ĸļ

-0.78

opol

-0.76

phia

-0.73

eming

-0.72

pha

-0.72

Ö¼

-0.71

ighed

-0.69

bole

-0.68

nothing

-0.67

dding

-0.66

POSITIVE LOGITS

 your

1.12

 yourself

0.91

 Your

0.91

 YOUR

0.84

 Yourself

0.79

 oneself

0.75

 custom

0.72

 homemade

0.72

 properly

0.71

 authentic

0.69

Activations Density 0.233%

phrases related to setting up and creating something in a technical or developmental context

instructions and actions related to setting up and developing software environments or applications

No Comments

No Known Activations