INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 mathemat

-0.67

awaru

-0.64

atics

-0.63

 fortun

-0.61

ATS

-0.61

 bere

-0.60

 millenn

-0.60

ongyang

-0.59

bling

-0.59

 immersion

-0.59

POSITIVE LOGITS

bucks

0.78

chin

0.75

sheet

0.71

ocument

0.68

crop

0.68

 Blueprint

0.67

\<

0.67

 Investor

0.66

enegger

0.66

 Carney

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.