INDEX

Explanations

concepts related to traits and characteristics in various contexts

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

describing characteristics or concepts

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

Autoritní

-0.58

 ainfi

-0.55

-0.54

OGND

-0.54

 ligiloj

-0.52

 Biôgrafia

-0.50

 protoimpl

-0.49

новниш

-0.49

 snippetHide

-0.48

 utafitiHapana

-0.48

POSITIVE LOGITS

 include

0.42

 Include

0.39

 صوتيه

0.38

 INCLUDE

0.37

 includes

0.33

include

0.33

 varies

0.32

initializeApp

0.32

нор

0.32

 antaranya

0.32

Activations Density 0.103%

concepts related to traits and characteristics in various contexts

describing characteristics or concepts

No Comments

No Known Activations