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

interstitial

-0.78

uble

-0.73

eatures

-0.71

ascript

-0.71

 elbows

-0.70

 Bulg

-0.68

 disputes

-0.64

 midrange

-0.63

omez

-0.62

 democratically

-0.62

POSITIVE LOGITS

Nat

0.67

Kot

0.66

TERN

0.65

 Plain

0.65

Station

0.64

 Sach

0.64

itialized

0.64

Site

0.63

Ort

0.63

 TEST

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.