INDEX

Explanations

No Explanations Found

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

ebin

-0.78

userc

-0.75

eland

-0.70

VK

-0.66

senal

-0.63

azo

-0.63

 complexity

-0.63

 goodwill

-0.61

WIN

-0.61

 mistress

-0.61

POSITIVE LOGITS

Newsletter

0.80

Accessory

0.76

alion

0.66

 Dull

0.66

ART

0.66

 Amend

0.65

Rail

0.65

Born

0.64

KT

0.63

 passers

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.