© 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. 2120
    Prev
    Next
    INDEX
    Explanations

    not uncommon

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

    The neuron sharply lights up on rare or unusually specific tokens—i.e. low‐frequency, out‐of‐vocabulary words.

    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
    an
    0.84
    er
    0.83
    a
    0.80
    on
    0.79
    ی
    0.73
    en
    0.71
    ers
    0.70
    서
    0.70
    ee
    0.67
    ment
    0.64
    POSITIVE LOGITS
    锈钢
    0.94
     hề
    0.94
     żad
    0.93
    丝毫
    0.82
     obstante
    0.80
    任何
    0.72
    啻
    0.72
     necessariamente
    0.72
     alcuna
    0.71
    безпе
    0.70
    Activations Density 0.000%

    No Known Activations