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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Virgin

-0.72

grave

-0.69

wr

-0.67

 ammon

-0.66

erer

-0.64

 Oval

-0.64

 pedals

-0.62

ery

-0.62

 slic

-0.61

ovi

-0.60

POSITIVE LOGITS

SPONSORED

0.81

soType

0.75

zig

0.72

Cth

0.71

 thous

0.68

tz

0.68

 predec

0.67

ONSORED

0.67

EMENT

0.67

 obser

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.