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

ItemImage

-0.79

printf

-0.74

 usher

-0.69

ItemTracker

-0.68

Reviewer

-0.68

 overrun

-0.68

âĢ¢âĢ¢

-0.67

catentry

-0.66

reply

-0.66

Enlarge

-0.63

POSITIVE LOGITS

 Unicorn

0.76

antine

0.68

angu

0.67

orns

0.67

ashes

0.66

ashed

0.64

 Dynasty

0.64

 Markets

0.64

OWER

0.63

ebra

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.