INDEX

Explanations

references to gratitude and appreciation

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

numbers and common words

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

PreAuthorize

-0.33

 productivity

-0.33

 resistance

-0.29

 security

-0.27

 initiatives

-0.27

 initiative

-0.27

borderWidth

-0.26

 Pernambuco

-0.26

 cardíaca

-0.25

POSITIVE LOGITS

 Paglinawan

0.67

出版年

0.65

 Taktlose

0.65

 nakalista

0.64

 betweenstory

0.62

 Administrativna

0.61

 Италијани

0.61



0.58

Abitanti

0.58

tanleria

0.58

Activations Density 0.001%

references to gratitude and appreciation

numbers and common words

No Comments

No Known Activations