INDEX

Explanations

statistical analysis terms and methods focused on comparisons between groups

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

post hoc tests

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

transQ

-0.54

Portály

-0.50

 itſelf

-0.50

endphp

-0.50

vaadin

-0.50

 незавершена

-0.49

 StatelessWidget

-0.49

テンツ

-0.48

 betweenstory

-0.47

crollView

-0.46

POSITIVE LOGITS

 Juga

0.52

 intStringLen

0.52

@[+][

0.47

 smtplib

0.47

ItemBackground

0.45

Controllo

0.44

aarrggbb

0.42

 Sequential

0.39

 beleid

0.38

Nationalité

0.36

Activations Density 0.010%

statistical analysis terms and methods focused on comparisons between groups

post hoc tests

No Comments

No Known Activations