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

utical

-0.74

reason

-0.67

 Interview

-0.65

 Mist

-0.63

cr

-0.62

 frank

-0.62

essions

-0.62

 honesty

-0.61

 nutshell

-0.61

 lucid

-0.60

POSITIVE LOGITS

 eleph

0.97

 Leban

0.92

MSN

0.82

 actionGroup

0.78

LET

0.73

FTWARE

0.72

PATH

0.69

yip

0.69

orius

0.68

 postage

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.