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

allery

-0.77

arre

-0.73

bars

-0.68

 Quarterly

-0.66

eatures

-0.66

river

-0.65

abeth

-0.64

 scapego

-0.64

escription

-0.64

ippi

-0.64

POSITIVE LOGITS

 days

1.14

Day

0.87

day

0.84

 mornings

0.77

DN

0.72

ially

0.71

Day

0.69

olin

0.68

days

0.67

 milliseconds

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.