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

geist

-0.71

eous

-0.65

 compass

-0.64

alian

-0.62

 afar

-0.61

umbn

-0.61

 Pride

-0.60

doi

-0.60

riad

-0.59

athom

-0.58

POSITIVE LOGITS

Downloadha

0.73

ipes

0.71

ibles

0.69

abilia

0.67

arial

0.66

 Canberra

0.65

ADS

0.65

oured

0.63

 Rohingya

0.62

''

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.