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

urtles

-0.88

 vain

-0.80

é¾įå

-0.80

obe

-0.73

xtap

-0.71

ambo

-0.71

irtual

-0.71

vez

-0.70

oros

-0.69

maxwell

-0.68

POSITIVE LOGITS

menu

0.74

ITION

0.67

ially

0.65

include

0.63

 PayPal

0.62

 Asus

0.61

 localization

0.61

 clipboard

0.60

SPA

0.60

FILE

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.