INDEX

Explanations

mentions of trying to understand or solve something

oai_token-act-pair · gpt-3.5-turbo

references to 'figure' in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

eatures

-0.79

bey

-0.78

bnb

-0.74

ruciating

-0.73

 Liberties

-0.73

 Rapids

-0.71

anus

-0.69

 mids

-0.69

herry

-0.68

itch

-0.67

POSITIVE LOGITS

 figure

1.14

figure

1.06

 figures

1.01

 Figures

0.94

 Figure

0.93

Figure

0.92

enance

0.91

book

0.83

fig

0.80

books

0.79

Activations Density 0.014%

mentions of trying to understand or solve something

references to 'figure' in various contexts

No Comments

No Known Activations