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

osphere

-0.84

ization

-0.80

vy

-0.79

izing

-0.76

ificent

-0.75

ieu

-0.75

irts

-0.75

oli

-0.74

olding

-0.73

tsky

-0.72

POSITIVE LOGITS

 veter

0.77

 conduc

0.76

 disabilities

0.76

Anim

0.75

 sclerosis

0.74

 compr

0.68

 eleph

0.65

 compet

0.62

 charact

0.62

Robin

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.