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

FIX

-0.79

\-

-0.72

><

-0.70

 Mane

-0.69

GD

-0.68

iour

-0.67

Player

-0.67

include

-0.64

ËĪ

-0.62

eful

-0.62

POSITIVE LOGITS

Peb

0.75

 mosqu

0.70

Faw

0.69

uple

0.68

attery

0.67

blu

0.66

llah

0.63

 Burgess

0.63

oub

0.63

 Baxter

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.