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

Ud

-0.75

 Dahl

-0.75

 redes

-0.71

 Venture

-0.71

 Halls

-0.70

 Bounty

-0.70

 Deer

-0.68

 Thro

-0.68

 Corpse

-0.67

 Odyssey

-0.67

POSITIVE LOGITS

Palest

0.74

gm

0.69

fecture

0.69

oire

0.64

zing

0.64

armed

0.64

 bless

0.63

haired

0.63

ibly

0.63

agher

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.