INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Shape

-0.91

ãĤ±

-0.72

ahu

-0.67

ãĤ¼ãĤ¦ãĤ¹

-0.66

ãĥ¼ãĥĨãĤ£

-0.65

 Bore

-0.64

ãģ®å®

-0.63

å¥³

-0.62

å¤©

-0.60

 Beng

-0.59

POSITIVE LOGITS

igree

0.86

asma

0.74

undy

0.70

imentary

0.69

rower

0.68

maxwell

0.67

 assorted

0.67

ilities

0.67

gerald

0.65

tered

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.