INDEX

Explanations

information related to official announcements and releases

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "official press release."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

MpServer

-0.85

cum

-0.83

¬¼

-0.82

Õ

-0.82

Years

-0.80

cigarettes

-0.74

severe

-0.74

seless

-0.73

assault

-0.71

ruciating

-0.70

POSITIVE LOGITS

 teaser

1.74

 announcement

1.54

 tease

1.41

 synopsis

1.39

 trailer

1.36

 promo

1.30

 announcements

1.29

 confirmation

1.28

 trailers

1.25

 screenshots

1.23

Activations Density 0.564%

information related to official announcements and releases

occurrences of the phrase "official press release."

No Comments

No Known Activations