© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 10211
    Prev
    Next
    INDEX
    Explanations

    The neuron is acting as a positional indicator, signaling how far into the sequence each token is by producing a high activation early on that steadily decays over the course of the input.

    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
    ObjectTemp
    0.77
    អេ
    0.74
    但不
    0.73
    Plaintiff
    0.73
    TempValue
    0.71
    ायतों
    0.71
    🌆
    0.70
    تاة
    0.69
    屯
    0.68
    🚱
    0.68
    POSITIVE LOGITS
    řed
    0.87
    vej
    0.84
    ız
    0.82
     této
    0.79
    rets
    0.77
     pří
    0.76
    ře
    0.75
    ás
    0.75
     ویس
    0.73
     곱
    0.71
    Activations Density 0.000%

    No Known Activations