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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mingham

-0.72

indal

-0.71

idity

-0.67

vati

-0.67

tes

-0.65

chip

-0.65

bage

-0.65

 sshd

-0.65

anwhile

-0.65

chio

-0.64

POSITIVE LOGITS

Â·Â·

0.69

 IMAGES

0.68

 duties

0.68

)'

0.64

 blame

0.64

 waivers

0.62

ction

0.60

 obligation

0.59

 disclaim

0.59

 Thumbnails

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.