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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çİĭ

-0.77

tem

-0.68

ãĥ¼

-0.65

 Corker

-0.64

maker

-0.60

vich

-0.60

ETF

-0.60

mia

-0.60

brance

-0.59

ãĥ¼ãĥ

-0.59

POSITIVE LOGITS

quished

0.71

 Entered

0.69

Ka

0.69

Flo

0.67

Flo

0.65

ounded

0.65

 Seasons

0.65

cha

0.65

rived

0.64

awed

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.