INDEX

Explanations

references to the TV show "Orange is the New Black."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Orange."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Ö¼

-0.98

sonian

-0.86

stood

-0.78

adr

-0.78

enance

-0.78

rets

-0.73

ngth

-0.73

akable

-0.72

arnaev

-0.71

ted

-0.70

POSITIVE LOGITS

vale

1.09

 peel

1.05

 Blossom

1.05

 juice

1.00

 Peel

0.97

 Juice

0.94

issance

0.89

 Crush

0.83

 cones

0.79

berry

0.78

Activations Density 0.018%

references to the TV show "Orange is the New Black."

occurrences of the word "Orange."

No Comments

No Known Activations