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

omial

-0.83

 Property

-0.78

tones

-0.72

à¥

-0.71

ogn

-0.69

 negro

-0.67

VID

-0.66

 Prim

-0.65

ampion

-0.64

 Armory

-0.63

POSITIVE LOGITS

sheets

0.82

zzle

0.74

sheet

0.72

CLOSE

0.71

FILE

0.70

OPLE

0.70

EE

0.68

rolley

0.67

Rev

0.67

uku

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.