INDEX

Explanations

cooking instructions or steps

oai_token-act-pair · gpt-3.5-turbo

instances of the word "cook" and its variations in a cooking context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ulous

-0.75

aurus

-0.69

itton

-0.69

ilitary

-0.67

asketball

-0.67

orb

-0.66

peria

-0.65

hesda

-0.65

 Territories

-0.63

pled

-0.63

POSITIVE LOGITS

stall

1.10

books

0.96

 cook

0.90

å£«

0.89

Cook

0.87

washer

0.87

 cooks

0.87

book

0.85

cook

0.81

 Bake

0.81

Activations Density 0.011%

cooking instructions or steps

instances of the word "cook" and its variations in a cooking context

No Comments

No Known Activations