INDEX

Explanations

mentions of media, interviews, or talk shows involving specific individuals

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

podcast hosts discussing

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

endphp

-0.49

 AssemblyCompany

-0.45

 تكبرها

-0.44

 nahilalakip

-0.40

TestingModule

-0.39

 Италијани

-0.39

 JpaRepository

-0.39

 Galería

-0.38

 DriverManager

-0.38

 PeEnEo

-0.38

POSITIVE LOGITS

 podcast

0.59

discus

0.54

podcast

0.53

findpost

0.52

 hosts

0.50

 discussing

0.50

Podcast

0.49

funk

0.49

hosts

0.48

 Hosts

0.48

Activations Density 0.036%

mentions of media, interviews, or talk shows involving specific individuals

podcast hosts discussing

No Comments

No Known Activations