INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Leban

-0.93

ĳå£«

-0.79

+#

-0.78

Disk

-0.72

 Clicker

-0.70

WithNo

-0.68

 srfAttach

-0.68

 Reincarn

-0.67

SPONSORED

-0.67

 "$:/

-0.66

POSITIVE LOGITS

att

0.75

nis

0.73

inary

0.72

aylor

0.70

hing

0.70

inates

0.69

any

0.69

hed

0.68

na

0.67

ibr

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.