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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

England

-0.72

ruciating

-0.70

 withd

-0.69

URI

-0.67

ãĥ¼ãĥĨ

-0.66

Liverpool

-0.66

 gall

-0.65

Sac

-0.64

Merit

-0.64

neys

-0.63

POSITIVE LOGITS

 himself

0.66

esthetic

0.63

0.62

 largeDownload

0.60

-->

0.60

ocations

0.60

hole

0.59

cape

0.59

>>

0.57

imation

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.