Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 2071
    Prev
    Next
    INDEX
    Explanations

    elements related to crime and the psychological impacts of obsession

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    associated with scary movies and stories

    np_acts-logits-general · gemini-2.0-flash

    horror and spooky movies

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    AccessorTable
    -0.70
     Picchu
    -0.54
    UnsafeEnabled
    -0.53
    zzlies
    -0.52
     <>",
    -0.52
     Rocket
    -0.51
     BorderRadius
    -0.50
    quiao
    -0.50
     administrativos
    -0.49
    rocket
    -0.49
    POSITIVE LOGITS
     horror
    1.30
     Horror
    1.26
    Horror
    1.19
    horror
    1.18
     Halloween
    1.15
     spooky
    1.09
     creepy
    1.03
    Halloween
    1.01
     scary
    1.00
     halloween
    0.99
    Activations Density 0.259%

    No Known Activations