© 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. 12591
    Prev
    Next
    INDEX
    Explanations

    3x3 cube, Tensorflow tf

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

    The neuron specifically detects occurrences of the relative pronoun “who.”

    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
    adanam
    0.70
    ivasena
    0.70
     chiar
    0.66
     били
    0.66
    牫
    0.64
     русских
    0.64
     გამოყენ
    0.63
     getAll
    0.62
    ürn
    0.62
    ករណ៍
    0.62
    POSITIVE LOGITS
     ease
    0.83
    依旧
    0.77
    Ease
    0.77
    駅
    0.77
     relive
    0.76
     incentivize
    0.74
    ক
    0.71
    その
    0.69
     revitalize
    0.68
     file
    0.68
    Activations Density 0.000%

    No Known Activations