INDEX

Explanations

symbolic or structural elements typically used in programming or markup languages

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

ſ character

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

+#+#

-0.84

GenerationType

-0.55

 BrowserModule

-0.53

Geplaatst

-0.53

ioutil

-0.52

밌

-0.52

enumii

-0.51

transQ

-0.50

GraphicsUnit

-0.49

󠁢

-0.49

POSITIVE LOGITS

 ſou

0.57

 ſen

0.55

 ſte

0.55

 ſever

0.54

 faſt

0.53

 itſelf

0.51

 tranſ

0.50

 eſſ

0.49

ſy

0.48

 neceſſ

0.48

Activations Density 0.018%

symbolic or structural elements typically used in programming or markup languages

ſ character

No Comments

No Known Activations