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

wcsstore

-0.73

oga

-0.70

ËĪ

-0.66

ãĤ¼ãĤ¦ãĤ¹

-0.65

Buzz

-0.64

baugh

-0.64

ho

-0.61

ixt

-0.61

kai

-0.61

vana

-0.60

POSITIVE LOGITS

advant

0.76

 impe

0.68

 mitigation

0.65

 Monteneg

0.63

 mitigating

0.62

 Ender

0.61

 Mald

0.61

defense

0.60

 Benghazi

0.59

 defense

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.