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

ModLoader

-0.97

minist

-0.65

 Diabetes

-0.65

iasis

-0.65

é¾įå¥ĳå£«

-0.64

ipment

-0.63

 charism

-0.62

EH

-0.61

ially

-0.61

 Whale

-0.60

POSITIVE LOGITS

ath

0.78

ahon

0.74

cheon

0.72

actionDate

0.70

anka

0.69

ats

0.68

 ballpark

0.67

boa

0.67

agle

0.65

amily

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.