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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Rosenberg

-0.72

 Merrill

-0.69

fam

-0.69

 Scher

-0.68

Myr

-0.68

McF

-0.68

 McDonnell

-0.68

ĳå£«

-0.67

 Picard

-0.67

mberg

-0.66

POSITIVE LOGITS

Interstitial

0.75

emale

0.74

hill

0.68

Lady

0.65

ï¸ı

0.65

Motion

0.64

UX

0.64

 vibration

0.64

 heck

0.64

guided

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.