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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MAT

-0.82

ãĤ¤ãĥĪ

-0.81

utra

-0.80

ħĭ

-0.80

ADRA

-0.78

metry

-0.78

æĪ¦

-0.76

hement

-0.76

RT

-0.75

imov

-0.74

POSITIVE LOGITS

vet

0.73

ilee

0.62

 Yankee

0.62

 Zoom

0.62

 unchecked

0.61

 insured

0.61

 Percentage

0.61

 backup

0.60

st

0.59

 Reserved

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.