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

orgetown

-0.76

ordan

-0.72

endars

-0.65

 Horowitz

-0.65

uries

-0.65

otics

-0.65

illion

-0.63

byn

-0.63

OTAL

-0.63

animous

-0.62

POSITIVE LOGITS

ILLE

0.76

ãĥĺ

0.67

staking

0.65

Lair

0.61

æĿ

0.61

æĹ

0.61

luc

0.60

 disguise

0.60

ãĤ¬

0.59

ãĥĥãĥĪ

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.