INDEX

Explanations

references to visual media and photographic content

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

phrases related to making requests or giving something

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

visual descriptions and observation

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Topics

-0.54

toHaveBeenCalled

-0.53

+:+

-0.51

 recall

-0.48

 eventName

-0.48

Assista

-0.47

 Eventos

-0.47

 endregion

-0.46

 turbin

-0.46

Points

-0.45

POSITIVE LOGITS

adaptiveStyles

0.83

providedIn

0.65

 kasarigan

0.64

 الحره

0.61

delwed

0.59

AnimationsModule

0.58

Nazionale

0.57

 représent

0.56

 незавершена

0.56

 encarga

0.54

Activations Density 0.289%

references to visual media and photographic content

phrases related to making requests or giving something

visual descriptions and observation

No Comments

No Known Activations