INDEX

Explanations

positive evaluations or features related to artistic or production quality in a creative work

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

start of turn user

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

verifyException

-0.84

 queſta

-0.79

transQ

-0.78

ftagPool

-0.77

 snippetHide

-0.77

ðsíða

-0.77

GEBURTSDATUM

-0.76

 laſſen

-0.75

الحياه

-0.72

 betweenstory

-0.72

POSITIVE LOGITS

0.29

 infatti

0.28

 donc

0.27

 mente

0.26

or

0.26

irs

0.24

 žád

0.23

 zelfs

0.23

 ciò

0.23

 geprägt

0.22

Activations Density 0.128%

positive evaluations or features related to artistic or production quality in a creative work

start of turn user

No Comments

No Known Activations