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

rack

-0.83

itsu

-0.74

acqu

-0.73

anqu

-0.72

rag

-0.71

analy

-0.70

cest

-0.69

phant

-0.68

wat

-0.67

amb

-0.67

POSITIVE LOGITS

å£«

1.10

 Syndicate

0.80

RANT

0.75

 Pigs

0.71

Ĥİ

0.71

 Mayhem

0.70

OPLE

0.67

 ANGEL

0.64

UID

0.62

 Cortex

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.