INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

insula

-0.76

velt

-0.75

å¾

-0.73

ç«

-0.72

ENG

-0.71

ICT

-0.68

 squared

-0.67

ILLE

-0.66

æŃ

-0.65

æĿ

-0.64

POSITIVE LOGITS

tons

0.90

 absent

0.80

 Beirut

0.69

rather

0.64

 alas

0.64

 acknowled

0.64

 Monaco

0.64

 nevertheless

0.63

ersen

0.62

 fortunately

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.