INDEX

Explanations

references to patent applications and publications

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

WO patent identifiers

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

$_['

-0.70

TagHelper

-0.57

الحياه

-0.54

xase

-0.54

 ModelExpression

-0.53

vdots

-0.52

 nonatomic

-0.51

wieś

-0.51

XDECREF

-0.51

UrlResolution

-0.50

POSITIVE LOGITS

endregion

0.48

 International

0.48

 international

0.48

 world

0.46

 internacional

0.46

 mundial

0.46

 mondial

0.45

世界

0.43

 worldwide

0.43

 internation

0.43

Activations Density 0.011%

references to patent applications and publications

WO patent identifiers

No Comments

No Known Activations