INDEX

Explanations

code elements related to function calls and object methods

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

zoom or light

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

 يتيمه

-0.74

 autorytatywna

-0.60

 queſta

-0.57

 propOrder

-0.57

 مرئيه

-0.57

WebElementEntity

-0.57

OGND

-0.54

richTextPanel

-0.52

 ujednoznacz

-0.52

featureID

-0.52

POSITIVE LOGITS

 also

0.69

 inoltre

0.54

are

0.45

 außerdem

0.45

 además

0.44

 également

0.44

 must

0.44

0.43

enregistrement

0.43

 controlled

0.42

Activations Density 0.021%

code elements related to function calls and object methods

zoom or light

No Comments

No Known Activations