© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1914
    Prev
    Next
    INDEX
    Explanations

    base, investment, installation, potential

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

    The neuron isn’t really looking for any particular words or concepts at all but is simply acting as a “position” detector—its activation grows (or shrinks) in near linear proportion to how far into the text you are.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    нове
    0.80
    ob
    0.75
    ায়ণ
    0.75
     grantees
    0.75
    ગે
    0.75
    㑔
    0.73
    sembles
    0.73
     будете
    0.71
     catastrophes
    0.71
    诒
    0.70
    POSITIVE LOGITS
     Bunu
    0.82
     Dzięki
    0.80
     però
    0.79
     Quotes
    0.77
     Fior
    0.76
     remarque
    0.74
     Međutim
    0.73
    .):
    0.73
     Tamaño
    0.72
     درصد
    0.71
    Activations Density 0.000%

    No Known Activations