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

eteria

-0.82

umbn

-0.69

assetsadobe

-0.69

onday

-0.67

 Reflex

-0.66

torn

-0.66

Express

-0.63

 beginner

-0.63

adapt

-0.63

keyes

-0.61

POSITIVE LOGITS

 unsub

0.66

adal

0.64

merce

0.64

HPV

0.64

chio

0.62

auga

0.61

 Tsukuyomi

0.61

ibe

0.60

 Lancet

0.59

 Vish

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.