INDEX

Explanations

the word "Okay" or similar expressions indicating agreement or approval

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Okay" or similar affirmations in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

advert

-0.77

 brim

-0.73

expression

-0.66

effic

-0.63

ocratic

-0.63

creation

-0.61

apers

-0.61

gaard

-0.59

repl

-0.58

clipse

-0.58

POSITIVE LOGITS

 Alright

1.16

Okay

1.14

Alright

1.10

lahoma

1.08

 Okay

1.06

bye

0.95

OK

0.91

 alright

0.91

 Guys

0.88

 okay

0.88

Activations Density 0.010%

the word "Okay" or similar expressions indicating agreement or approval

instances of the word "Okay" or similar affirmations in the text

No Comments

No Known Activations