INDEX

Explanations

references to humor or things perceived as funny

oai_token-act-pair · gpt-3.5-turbo

references to humor or things considered funny

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

ainer

-0.95

IENT

-0.83

irlwind

-0.81

ilings

-0.79

hips

-0.78

avored

-0.75

ighth

-0.74

aer

-0.74

hip

-0.73

eding

-0.72

POSITIVE LOGITS

 funny

1.09

 Funny

0.81

 Haram

0.79

 Comed

0.77

Pwr

0.75

 enough

0.73

 predic

0.72

balls

0.71

 quirks

0.71

 comedy

0.70

Activations Density 0.007%

references to humor or things perceived as funny

references to humor or things considered funny

No Comments

No Known Activations