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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

KN

-0.73

KNOWN

-0.73

ãĥİ

-0.68

abbage

-0.68

ministic

-0.67

photo

-0.66

ä½¿

-0.64

EY

-0.63

BOOK

-0.63

 swearing

-0.63

POSITIVE LOGITS

hens

0.67

 Rockies

0.66

itcher

0.65

MLB

0.63

NetMessage

0.61

Iz

0.60

river

0.60

 Gomez

0.60

 Chimera

0.60

 Griffin

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.