INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

meier

-0.68

tein

-0.66

jing

-0.65

container

-0.62

elong

-0.62

riber

-0.61

oglu

-0.61

 basin

-0.58

 Bearing

-0.58

heimer

-0.58

POSITIVE LOGITS

 DirectX

0.81

Goo

0.77

DX

0.74

PHOTOS

0.69

uild

0.69

LI

0.66

HDL

0.66

 McAuliffe

0.65

.}

0.65

ãĥ¯

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.