INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

teenth

-0.78

INGS

-0.76

eting

-0.71

ISTER

-0.70

Ô

-0.68

 unaccompanied

-0.68

Det

-0.65

à¼

-0.65

ENN

-0.65

âĺħâĺħ

-0.65

POSITIVE LOGITS

swick

0.74

yang

0.67

communications

0.65

VP

0.64

 Blazers

0.62

yrinth

0.62

mite

0.61

hub

0.60

oliath

0.59

 cous

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.