INDEX

Explanations

technical terms related to unauthorized access or intrusion

oai_token-act-pair · gpt-3.5-turbo

terms related to intrusion or encroachment

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

creen

-0.91

fully

-0.84

hift

-0.78

baum

-0.77

FUL

-0.75

wagen

-0.73

LESS

-0.73

hare

-0.72

fulness

-0.70

rique

-0.68

POSITIVE LOGITS

uders

1.24

uding

1.11

usions

1.00

ordinary

0.99

asonic

0.99

usted

0.96

aband

0.92

udes

0.91

acent

0.91

inx

0.89

Activations Density 0.046%

technical terms related to unauthorized access or intrusion

terms related to intrusion or encroachment

No Comments

No Known Activations