INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ifacts

-0.80

æ©Ł

-0.67

Quality

-0.64

pedia

-0.63

worthy

-0.63

 Minion

-0.62

å¸

-0.60

 Hero

-0.60

 Detail

-0.59

................................................................

-0.59

POSITIVE LOGITS

ovie

0.77

anza

0.77

atch

0.74

liter

0.73

exit

0.72

olt

0.71

 disgust

0.71

udeb

0.69

pez

0.69

otin

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.