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

vernment

-0.73

 Hust

-0.68

SUP

-0.64

ragon

-0.63

GROUND

-0.62

Rock

-0.62

 Sneak

-0.62

eds

-0.62

 sling

-0.61

 VIDEOS

-0.61

POSITIVE LOGITS

ng

0.75

vre

0.71

ulet

0.68

umpy

0.67

nton

0.67

achus

0.65

ulus

0.65

ubb

0.64

".

0.63

oke

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.