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

undreds

-0.71

BUG

-0.70

ICLE

-0.69

 nodd

-0.68

ixties

-0.65

ogue

-0.64

achev

-0.63

gin

-0.62

 braces

-0.61

IDES

-0.60

POSITIVE LOGITS

chance

0.78

 Chun

0.67

akia

0.65

miss

0.62

 Hels

0.60

 Motorsport

0.58

 Invaders

0.57

wed

0.56

rious

0.56

eno

0.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.