INDEX

Explanations

high activation values indicating successful processing or responses in a structured programming context

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

code closing braces

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ffindor

-0.48

Демографія

-0.45

enterOuterAlt

-0.41

Referencie

-0.39

riotic

-0.38

posx

-0.38

 integral

-0.37

 Brotherhood

-0.37

cool

-0.36

 zentralen

-0.36

POSITIVE LOGITS

 مرئيه

0.54

 صوتيه

0.53

 <=",

0.50

 الحره

0.44

 تضيفلها

0.44

ьаж

0.43

ISupport

0.43

脚注の使い方

0.42

 мәкал

0.42

enderror

0.41

Activations Density 0.004%

high activation values indicating successful processing or responses in a structured programming context

code closing braces

No Comments

No Known Activations