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

ROR

-0.62

inton

-0.62

 ISBN

-0.61

onis

-0.61

borgh

-0.59

 Franch

-0.59

 vulnerabilities

-0.59

enz

-0.58

 multipl

-0.58

 Rout

-0.58

POSITIVE LOGITS

ional

0.74

Saharan

0.73

 whine

0.70

NG

0.68

imester

0.67

abee

0.67

aptic

0.66

ounced

0.66

ioned

0.65

eworthy

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.