INDEX

Explanations

references to the streaming service Netflix

oai_token-act-pair · gpt-3.5-turbo

mentions of Netflix

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

area

-0.72

cal

-0.70

upon

-0.65

idates

-0.63

Chem

-0.62

XII

-0.62

Place

-0.61

vae

-0.60

fort

-0.60

 pads

-0.60

POSITIVE LOGITS

 Netflix

3.75

Netflix

3.43

 Hulu

2.34

netflix

2.29

flix

1.92

HBO

1.84

 Roku

1.79

 Spotify

1.76

 Showtime

1.65

 Airbnb

1.60

Activations Density 0.019%

references to the streaming service Netflix

mentions of Netflix

No Comments

No Known Activations