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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

breaks

-0.77

rez

-0.72

Ez

-0.70

anke

-0.67

gradation

-0.66

 jams

-0.66

vez

-0.63

olis

-0.62

enhagen

-0.62

bling

-0.61

POSITIVE LOGITS

 Merchants

0.79

çİĭ

0.74

ONY

0.73

ħĭ

0.72

 Users

0.70

estamp

0.69

¬¼

0.68

CENT

0.68

 Patron

0.66

EFF

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.