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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 7620
    Prev
    Next
    INDEX
    Explanations

    code snippets and technical formats

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

    This neuron primarily responds to the small function word “to,” especially when it appears as the separator in date or number ranges.

    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
     isoforms
    0.76
    اعت
    0.74
     İlk
    0.73
     Đối
    0.73
    ר
    0.72
    ర్లు
    0.71
    ν
    0.70
    עים
    0.69
    vée
    0.69
    רת
    0.67
    POSITIVE LOGITS
    ton
    0.84
    Boogie
    0.80
    спомина
    0.79
    ."))
    0.78
     sogno
    0.78
    .${
    0.77
     spurious
    0.77
     conexão
    0.77
    $">
    0.77
     сообщает
    0.76
    Activations Density 0.027%

    No Known Activations