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

isode

-0.73

 Franch

-0.65

ories

-0.63

iar

-0.62

ãĥ¼ãĥĨãĤ£

-0.61

 Mesa

-0.60

cli

-0.59

ãĥ´ãĤ¡

-0.59

itions

-0.59

insula

-0.58

POSITIVE LOGITS

\/\/

0.75

bish

0.67

brush

0.66

acco

0.65

acht

0.64

\\\\\\\\

0.64

ITNESS

0.62

_>

0.62

Äĩ

0.62

uggle

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.