INDEX

Explanations

instances of the word "supposedly"

oai_token-act-pair · gpt-3.5-turbo

the word "supposedly" indicating statements that imply doubt or uncertainty

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

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

cair

-0.72

ger

-0.71

 Dragonbound

-0.71

 hearts

-0.70

 Roof

-0.67

externalActionCode

-0.66

 Clicker

-0.66

oft

-0.64

aughs

-0.64

 heart

-0.64

POSITIVE LOGITS

æ©

0.78

 metic

0.73

 explan

0.71

ãĥ¼ãĥĨãĤ£

0.69

Buyable

0.66

 incent

0.65

ãĤ´ãĥ³

0.65

ĪĴ

0.64

 contradict

0.63

mas

0.63

Activations Density 0.008%

instances of the word "supposedly"

the word "supposedly" indicating statements that imply doubt or uncertainty

No Comments

No Known Activations