© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1648
    Prev
    Next
    INDEX
    Explanations

    Job titles and executive positions

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

    The neuron strongly activates on words denoting high‐level professional roles or titles (e.g. vice, president, director, CEO).

    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_31_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
     Sebagai
    0.55
     作为
    0.52
    作为
    0.52
     sebagai
    0.50
    作为一个
    0.50
     Министер
    0.48
    毎
    0.48
    律师
    0.47
    作為
    0.46
     Як
    0.45
    POSITIVE LOGITS
     emeritus
    1.00
     Emeritus
    0.76
    兼
    0.63
    /
    0.61
     overseeing
    0.59
     designee
    0.59
     extraordinaire
    0.57
     designate
    0.57
     ofthe
    0.54
     thereof
    0.51
    Activations Density 0.226%

    No Known Activations