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

 skelet

-0.70

 toile

-0.68

nutrition

-0.66

daq

-0.63

Dav

-0.61

 naked

-0.61

Nib

-0.61

ga

-0.61

"]=>

-0.61

stro

-0.60

POSITIVE LOGITS

ahime

0.79

ãĤ¬

0.76

ARP

0.70

iliar

0.68

iple

0.68

hunt

0.67

pled

0.66

ursions

0.66

piracy

0.63

edia

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.