INDEX

Explanations

the word "not" followed by a verb

oai_token-act-pair · gpt-3.5-turbo

negations or phrases indicating refusal or denial

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

krit

-0.62

ilo

-0.55

ixel

-0.53

 Disk

-0.51

pot

-0.51

 Valhalla

-0.51

tein

-0.50

 Tycoon

-0.50

CSS

-0.49

idon

-0.48

POSITIVE LOGITS

epad

1.02

ices

0.92

ifying

0.92

ifies

0.90

to

0.90

hin

0.90

ifier

0.88

ifications

0.86

ification

0.83

ifiers

0.82

Activations Density 0.033%

the word "not" followed by a verb

negations or phrases indicating refusal or denial

No Comments

No Known Activations