INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Interstitial

-0.75

enegger

-0.67

 Patient

-0.63

 Deadpool

-0.63

 Brands

-0.63

 Gentle

-0.60

 Nurse

-0.59

 Petr

-0.59

 pharmacies

-0.58

 Quiet

-0.58

POSITIVE LOGITS

sonian

0.92

bol

0.82

inctions

0.74

dayName

0.73

inguished

0.72

ional

0.72

erey

0.72

inction

0.71

lin

0.70

aleb

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.