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

 Thumbnails

-0.78

arch

-0.68

oshop

-0.67

quickShipAvailable

-0.67

oler

-0.67

ospace

-0.66

iven

-0.64

hemer

-0.64

reet

-0.63

ĸļ

-0.62

POSITIVE LOGITS

 Celt

0.73

 Malfoy

0.65

enough

0.63

asca

0.62

AMD

0.62

fast

0.61

Yan

0.61

ãĥ¼ãĥĨãĤ£

0.61

 McCull

0.60

crop

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.