INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

asio

-0.73

marg

-0.68

pless

-0.67

gettable

-0.66

ereo

-0.65

them

-0.64

etheless

-0.64

rete

-0.63

lag

-0.62

Sport

-0.62

POSITIVE LOGITS

enegger

0.73

CAST

0.67

 Trans

0.65

 Adin

0.63

 sidx

0.63

é¾įå¥ĳå£«

0.62

ENDED

0.62

 tatt

0.61

é»Ĵ

0.60

RED

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.