INDEX

Explanations

numeric values, particularly in lists or structured data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_15/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 Infór

-0.54

 ویکی‌پدیا

-0.52

 Мексичка

-0.52

 pétrole

-0.52

 kentang

-0.51

 hasilnya

-0.50

 Económica

-0.50

 abstrato

-0.49

 vectorielle

-0.49

 pengunjung

-0.49

POSITIVE LOGITS

iomanip

0.48

ACHUSET

0.48

 aligned

0.48

cellence

0.47

 Align

0.47

ChildScrollView

0.46

AJ

0.46

 chal

0.46

 commissioning

0.45

 Chron

0.45

Activations Density 0.614%

numeric values, particularly in lists or structured data

No Comments

No Known Activations