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

agra

-0.81

akuya

-0.77

 Defin

-0.71

eny

-0.67

arin

-0.66

anship

-0.66

thood

-0.65

ilan

-0.65

soDeliveryDate

-0.62

agame

-0.61

POSITIVE LOGITS

Corp

0.76

Decre

0.73

grad

0.72

INGTON

0.66

ixels

0.65

antam

0.65

CNN

0.63

malink

0.63

produ

0.62

batch

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.