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

llah

-0.85

roph

-0.68

ople

-0.68

Pear

-0.68

tery

-0.67

¶ħ

-0.66

ħĭ

-0.66

pload

-0.65

axter

-0.65

"></

-0.64

POSITIVE LOGITS

alys

0.68

 GOODMAN

0.66

_-

0.64

english

0.63

 kilometres

0.63

imedia

0.62

æ©

0.61

 kilometers

0.61

 directs

0.60

 standings

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.