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

+++

-0.83

 Firefly

-0.69

heads

-0.68

work

-0.67

books

-0.64

 ROCK

-0.64

,—

-0.63

bane

-0.63

++++

-0.62

bottom

-0.62

POSITIVE LOGITS

olson

0.78

imet

0.69

andel

0.66

abet

0.65

irtual

0.65

itchie

0.64

ults

0.62

avanaugh

0.62

amation

0.62

akespe

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.