INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Orient

-0.76

oshenko

-0.70

 incom

-0.67

 digit

-0.66

cone

-0.63

htaking

-0.63

 saga

-0.63

urus

-0.62

 averaging

-0.62

 hemisphere

-0.62

POSITIVE LOGITS

emis

0.84

assetsadobe

0.77

Interstitial

0.75

etus

0.74

ãĤ¡

0.73

lasses

0.71

LECT

0.65

STD

0.65

obiles

0.64

soType

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.