INDEX

Explanations

instances of the first person speaking about an action they will do soon

oai_token-act-pair · gemini-2.0-flash

prefixes indicating future actions or posts

np_token-act-pair-logits · gpt-4o-mini

Future actions or plans

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 سكانية

-0.88

 Waray

-0.70

 الحره

-0.68

 ProtoMessage

-0.68

 الرياضيه

-0.68

RectangleBorder

-0.66

Читати

-0.65

 diyor

-0.65

/***/

-0.65

 AssemblyVersion

-0.65

POSITIVE LOGITS

 post

0.60

mod

0.45

ГЛА

0.43

 mañana

0.43

塞

0.42

 clean

0.42

 peut

0.42

saraba

0.41

 posts

0.41

hoog

0.40

Activations Density 2.072%

instances of the first person speaking about an action they will do soon

prefixes indicating future actions or posts

Future actions or plans

No Comments

No Known Activations