INDEX

Explanations

names of individuals

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

names followed by end markers

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

WebElementEntity

-0.56

DIPSETTING

-0.47

bildēt

-0.43

 siihen

-0.43

 daarna

-0.42

UserScript

-0.42

 Vielzahl

-0.42

mybatisplus

-0.41

 Berührung

-0.41

 arvio

-0.41

POSITIVE LOGITS

RefNanny

0.45

rungsseite

0.42

awtextra

0.41

transQ

0.41

 purpoſe

0.39

buttonBar

0.39

पया

0.38

 beſt

0.38

 folks

0.38

})));

0.38

Activations Density 0.013%

names of individuals

names followed by end markers

No Comments

No Known Activations

names of individuals

names followed by end markers

No Comments

No Known Activations