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

earances

-0.80

ower

-0.67

owers

-0.66

earance

-0.66

kaya

-0.65

board

-0.63

onew

-0.63

antam

-0.62

 Harbour

-0.62

lets

-0.62

POSITIVE LOGITS

lov

0.75

DOS

0.72

LOG

0.72

FY

0.67

UGE

0.67

Russ

0.64

successful

0.63

âķĲ

0.63

Ü

0.62

ACH

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.