INDEX

Explanations

constructs related to programming and data manipulation

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

next cursor

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

_("

-0.40

!("{}",

-0.40

setcounter

-0.39

 Италијани

-0.39

ScopeManager

-0.38

 strdup

-0.38

endforeach

-0.38

Aholisi

-0.37

kwds

-0.37

insee

-0.36

POSITIVE LOGITS

httphttps

0.60

AsUp

0.54

ConstraintMaker

0.54

 InputDecoration

0.50

routeProvider

0.49

 ویکی‌پدیای

0.48

AndEndTag

0.48

KommentareTeilen

0.48

GraphicsUnit

0.47

ⓧ

0.47

Activations Density 0.019%

constructs related to programming and data manipulation

next cursor

No Comments

No Known Activations