INDEX

Explanations

complex mathematical or computational terms and structures

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

rank of, row of, identity matrix

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

 دیکھیے

-0.49

ßerdem

-0.49

ContentAsync

-0.44

OGND

-0.44

الإنجليزية

-0.44

vocable

-0.44

 nahilalakip

-0.43

ipedi

-0.42

Sei

-0.42

 censiti

-0.42

POSITIVE LOGITS

Filmographie

0.54

vierno

0.49

webElementXpaths

0.45

 fermé

0.40

 faudra

0.39

 surprised

0.38

 jefes

0.38

 liderança

0.38

Rüyada

0.38

 sorprendió

0.38

Activations Density 0.055%

complex mathematical or computational terms and structures

rank of, row of, identity matrix

No Comments

No Known Activations