INDEX

Explanations

numerical values or metrics related to a context, possibly indicating data or statistical significance

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

phone numbers and email addresses

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

ſelf

-0.46

XmlAccessType

-0.46

">//

-0.44

<caption>

-0.43

 coelho

-0.41

 preſ

-0.41

 BrowserModule

-0.41

 AspNetCore

-0.41

 anſ

-0.40

 morango

-0.40

POSITIVE LOGITS

 IBOutlet

0.53

 CanadaChoose

0.46

 PHONE

0.41

nezeu

0.40

 vigencia

0.39

 emergencia

0.38

 الحره

0.37

ANSA

0.37

Tel

0.37

 atención

0.36

Activations Density 0.012%

numerical values or metrics related to a context, possibly indicating data or statistical significance

phone numbers and email addresses

No Comments

No Known Activations