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

Ń·

-0.71

¬¼

-0.70

sold

-0.70

 liner

-0.69

 folding

-0.64

iHUD

-0.63

ilde

-0.62

entimes

-0.60

Sov

-0.60

 redeemed

-0.58

POSITIVE LOGITS

machine

0.68

 Rodham

0.67

rer

0.67

enge

0.65

 fingert

0.65

vasive

0.64

 GOODMAN

0.62

gdala

0.62

axy

0.62

gency

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.