INDEX

Explanations

phrases involving inquiries for confirmation or information

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

know if what

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

Predecesor

-0.40

 médecins

-0.38

 BoxFit

-0.38

<?

-0.37

 defaultdict

-0.36

XSSF

-0.35

%"),

-0.35

glLoad

-0.34

 egip

-0.34

 fibre

-0.34

POSITIVE LOGITS

+#+#

0.62

 Winslow

0.57

omitempty

0.56

ɚ

0.55

 interested

0.54

 khảo

0.53

{}{}

0.52

tagHelperRunner

0.52

 initComponents

0.52

 expres

0.52

Activations Density 0.004%

phrases involving inquiries for confirmation or information

know if what

No Comments

No Known Activations