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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ogens

-0.76

igmatic

-0.75

eyes

-0.69

é¾įå¥ĳå£«

-0.69

rowing

-0.67

 Saunders

-0.66

wcs

-0.65

uckles

-0.65

terness

-0.64

imposed

-0.64

POSITIVE LOGITS

estamp

0.68

rants

0.64

 Diesel

0.63

 Hail

0.62

weed

0.60

ularity

0.60

 Trailer

0.60

 Pione

0.59

 Liter

0.59

Die

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.