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

ractical

-0.72

adle

-0.68

cs

-0.63

icut

-0.63

 Bosnia

-0.63

shed

-0.62

ctors

-0.62

ONSORED

-0.62

ivia

-0.62

fully

-0.62

POSITIVE LOGITS

ortium

0.85

\":

0.70

 procedural

0.69

\<

0.67

 Duration

0.67

arious

0.66

ularity

0.66

angles

0.64

accompan

0.63

gravity

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.