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

 minced

-0.64

cussion

-0.63

 racks

-0.63

acters

-0.62

 Pric

-0.60

 throats

-0.60

 registry

-0.59

venants

-0.59

 diced

-0.59

zig

-0.58

POSITIVE LOGITS

orph

0.69

 Eater

0.69

eton

0.68

egal

0.68

é¾įåĸļå£«

0.66

urst

0.66

Bey

0.65

essential

0.65

azon

0.64

uner

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.