INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

lem

-0.85

legraph

-0.70

birth

-0.69

raint

-0.68

SIGN

-0.66

alli

-0.66

ordered

-0.65

legram

-0.64

enz

-0.64

cash

-0.64

POSITIVE LOGITS

 fishes

0.73

livious

0.67

Þ

0.67

 Duchess

0.66

 dehyd

0.65

 mint

0.64

utral

0.64

 ducks

0.63

 mounts

0.62

OPA

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.