INDEX

Explanations

specific technical terms and nomenclature used in scientific contexts

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

mathematical and code 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

AISSEE

-0.53

VYMaps

-0.51

$_['

-0.39

WaitGroup

-0.37

astéroïdes

-0.37

 ragioni

-0.37

 tramonto

-0.37

 őket

-0.37

divers

-0.36

exemple

-0.36

POSITIVE LOGITS

tagHelperRunner

0.50

RegistryLite

0.47

 gabbana

0.44

Lähteet

0.42

 تضيفلها

0.41

 CanadaChoose

0.40

setopt

0.39

一代

0.38

 الأحمر

0.37

 править

0.37

Activations Density 0.213%

specific technical terms and nomenclature used in scientific contexts

mathematical and code identifiers

No Comments

No Known Activations