INDEX

Explanations

numerical data related to sales and performance metrics

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

initial states

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

HasMaxLength

-0.41

 licencia

-0.41

ellido

-0.41

ฉัน

-0.37

 ComVisible

-0.37

anskje

-0.35

 löytyy

-0.35

 aéreo

-0.34

verifyException

-0.33

 Paese

-0.32

POSITIVE LOGITS

 initial

0.55

⤒

0.53

 flooded

0.53

うちに

0.53

 INITIAL

0.53

 Initial

0.52

itial

0.49

mple

0.49

initial

0.48

avit

0.48

Activations Density 0.011%

numerical data related to sales and performance metrics

initial states

No Comments

No Known Activations