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

 secrecy

-0.77

ANN

-0.69

EB

-0.68

 XCOM

-0.68

BET

-0.66

 Byte

-0.66

ANG

-0.66

 Scientology

-0.63

 Tactics

-0.62

 Integrity

-0.61

POSITIVE LOGITS

ards

0.80

ister

0.79

isters

0.79

ensed

0.75

quished

0.74

urred

0.73

ibrarian

0.73

idth

0.72

jamin

0.71

nih

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.