INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

MAT

-0.71

eto

-0.68

ò

-0.66

ô

-0.65

 framed

-0.62

è¦ļéĨĴ

-0.61

vy

-0.61

 ALEC

-0.61

 McDonnell

-0.61

 helicop

-0.61

POSITIVE LOGITS

isu

0.77

itsu

0.69

 favourites

0.68

isations

0.67

Scroll

0.67

 quarrel

0.63

properties

0.62

appers

0.62

iere

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.