INDEX

Explanations

phrases indicating specific time references and sequences

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

"in" followed by definite/indefinite article

np_acts-logits-general · gemini-2.0-flash

late in or early in

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_16k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 dépass

-0.36

 juridiques

-0.35

 économiques

-0.34

Bibliograf

-0.34

 personnelles

-0.34

WriteAttribute

-0.34

çoivent

-0.33

 privées

-0.32

 uživ

-0.32

 PERMAN

-0.32

POSITIVE LOGITS

 throughout

0.71

 Throughout

0.67

Throughout

0.67

throughout

0.65

AtEndOf

0.64

 disambiguazione

0.60

 للمعارف

0.59

 ходу

0.59

endphp

0.57

 виправивши

0.57

Activations Density 0.408%

phrases indicating specific time references and sequences

"in" followed by definite/indefinite article

late in or early in

No Comments

No Known Activations