© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 9163
    Prev
    Next
    INDEX
    Explanations

    code punctuation

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

    the neuron detects direct‐address or call‐to‐action language—i.e. imperatives and reader‐directed phrasing featuring “you” and “to.”

    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
     eigenstates
    0.86
     сегодняш
    0.85
    fdPar
    0.84
     красоты
    0.83
    噦
    0.82
     Carpath
    0.81
    Сегодня
    0.81
     OCPP
    0.81
    لل
    0.81
    ڀ
    0.81
    POSITIVE LOGITS
    ism
    0.80
    0.80
    ps
    0.79
    ""
    0.75
     there
    0.73
    il
    0.73
    ts
    0.72
    ence
    0.71
    there
    0.71
    帷
    0.71
    Activations Density 0.000%

    No Known Activations