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

ildo

-0.77

 dime

-0.70

opa

-0.69

otherapy

-0.69

 princess

-0.68

OTUS

-0.67

 HERO

-0.65

 Genie

-0.63

 Ange

-0.63

uca

-0.63

POSITIVE LOGITS

stown

0.70

 Leth

0.66

ffield

0.65

anooga

0.64

 dism

0.63

antle

0.63

Smy

0.63

leton

0.62

erville

0.61

ahan

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.