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

ãĥ¼ãĥĨ

-0.84

Merit

-0.75

ogh

-0.70

 Barnett

-0.66

 gobl

-0.65

 sidx

-0.63

 ingred

-0.62

endez

-0.62

ãĤ¡

-0.62

 comprom

-0.62

POSITIVE LOGITS

alde

0.73

Introduced

0.71

bound

0.67

aired

0.67

ades

0.65

ija

0.65

ardo

0.64

emption

0.63

aura

0.62

auri

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.