INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Chicken

-0.71

hani

-0.71

amn

-0.71

cloth

-0.69

atche

-0.68

mud

-0.66

agine

-0.66

£ı

-0.66

arts

-0.65

oodle

-0.65

POSITIVE LOGITS

 targ

0.72

 weddings

0.67

ividual

0.67

Cov

0.66

 Liberties

0.66

versions

0.66

 homeowner

0.66

 Lerner

0.64

benef

0.64

 Raid

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.