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

ista

-0.69

soDeliveryDate

-0.68

istas

-0.67

Congress

-0.65

jay

-0.65

PokÃ©

-0.65

reci

-0.64

lez

-0.62

doi

-0.61

lain

-0.61

POSITIVE LOGITS

ached

0.76

 Rune

0.72

asse

0.67

AX

0.66

igious

0.66

 Abyss

0.64

chwitz

0.64

kaya

0.62

 Rath

0.62

iox

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.