INDEX

Explanations

specialized terminology from scientific and technical contexts

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

exchange field

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

Билгалдахарш

-0.73

 nahilalakip

-0.59

 InputDecoration

-0.50

 jsPsych

-0.49

 BoxDecoration

-0.49

 houſe

-0.46

脚注の使い方

-0.46

 Houſe

-0.46

 enfans

-0.45

WEBPACK

-0.45

POSITIVE LOGITS

ConverterFactory

0.55

contentLoaded

0.54

RepeatedField

0.54

 препратки

0.52

Personendaten

0.52

autoreleasepool

0.49

 iNdEx

0.49

InstrumentedTest

0.49

 Италијани

0.47

dafx

0.47

Activations Density 0.148%

specialized terminology from scientific and technical contexts

exchange field

No Comments

No Known Activations