INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

cats

-0.80

unic

-0.71

 bedrooms

-0.71

ecast

-0.66

etooth

-0.66

umber

-0.64

 Cust

-0.64

tu

-0.63

 Password

-0.63

wow

-0.62

POSITIVE LOGITS

DAY

0.69

Haunted

0.66

rived

0.64

Courtesy

0.64

SX

0.64

terson

0.62

andr

0.62

tnc

0.62

ilated

0.60

izable

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.