INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Premium

-0.74

ONSORED

-0.67

 Huckabee

-0.66

 PsyNet

-0.66

itent

-0.66

 Canaver

-0.65

overs

-0.64

uphem

-0.63

 Corker

-0.63

Teg

-0.63

POSITIVE LOGITS

£ı

0.76

»Ĵ

0.75

dot

0.73

Ń·

0.70

MpServer

0.67

brance

0.66

vice

0.65

 mars

0.65

bane

0.64

 Matter

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.