INDEX

Explanations

phrases related to educational materials and resources

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

provides description or resource

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

/**

-0.65

 linkovi

-0.57

舺

-0.56

 SafeMath

-0.56

//

-0.54

 betweenstory

-0.51

Gambas

-0.50

 Biôgrafia

-0.49

 تضيفلها

-0.49

 Мексичка

-0.49

POSITIVE LOGITS

 valuable

0.40

standen

0.39

proz

0.38

 TestBed

0.35

 implementation

0.35

arroll

0.34

 valioso

0.34

MESSAGES

0.33

 breakthrough

0.33

exitRule

0.33

Activations Density 0.006%

phrases related to educational materials and resources

provides description or resource

No Comments

No Known Activations