INDEX

Explanations

code-related structural elements and references in programming contexts

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

compound word formation after common prefixes

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

 iconLine

-0.56

󠁿

-0.56

 dAtA

-0.54

 مشين

-0.53

\{\\

-0.53

 Económica

-0.52

 فريبيس

-0.52

 invokingState

-0.51

 zelve

-0.51

󠁢

-0.50

POSITIVE LOGITS

oredCriteria

0.56

>//

0.55

----</

0.53

 ویکی‌پدی

0.52

 nonUne

0.50

},[])

0.44

IonicModule

0.44

 autorytatywna

0.44

angliski

0.43

pushFollow

0.43

Activations Density 1.594%

code-related structural elements and references in programming contexts

compound word formation after common prefixes

No Comments

No Known Activations