INDEX

Explanations

mentions of appearances or features in different media

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in" and variations of the phrase structure involving appearances in contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 teamwork

-0.68

 assured

-0.66

 maintained

-0.64

 persever

-0.64

 accommod

-0.63

 respecting

-0.62

.–

-0.62

 mortg

-0.62

 manipulation

-0.62

 mishand

-0.61

POSITIVE LOGITS

Picture

0.91

OPLE

0.81

¥µ

0.74

 Cele

0.73

 Playboy

0.72

DAQ

0.71

ARA

0.71

DragonMagazine

0.70

 Occupations

0.70

ivals

0.69

Activations Density 0.177%

mentions of appearances or features in different media

occurrences of the word "in" and variations of the phrase structure involving appearances in contexts

No Comments

No Known Activations