INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

geant

-0.65

âĶĢâĶĢ

-0.65

pent

-0.61

cise

-0.60

bly

-0.60

payment

-0.60

portion

-0.60

Vision

-0.60

lled

-0.58

venth

-0.58

POSITIVE LOGITS

airo

0.74

Sov

0.72

Sov

0.70

ed

0.67

RSS

0.67

usalem

0.66

ovych

0.65

istries

0.65

AFB

0.63

 paed

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.