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

 rematch

-0.72

BN

-0.70

zn

-0.66

license

-0.65

beat

-0.64

bid

-0.64

 defer

-0.62

reddit

-0.62

cdn

-0.61

Ub

-0.61

POSITIVE LOGITS

ãĥĥãĥī

0.77

accompanied

0.68

 Canaver

0.67

 Seah

0.66

CoC

0.63

ãĥĺãĥ©

0.62

urized

0.62

ocated

0.61

enhagen

0.61

 Alto

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.