INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

dark

-0.83

vo

-0.79

pie

-0.78

odus

-0.77

arenthood

-0.74

cube

-0.73

oat

-0.73

inez

-0.71

bott

-0.71

drug

-0.70

POSITIVE LOGITS

 Niger

0.81

 Adin

0.80

 Falls

0.74

 Anat

0.69

 Thib

0.69

 Nigeria

0.68

 Associ

0.68

 Constantin

0.68

 UNESCO

0.67

 GÃ¶

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.