© 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. 12236
    Prev
    Next
    INDEX
    Explanations

    base64 encoded strings

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

    The neuron detects superlative query phrases, especially “what is the best” in headings or questions.

    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
     🔥
    0.56
     کیا
    0.56
     Gallardo
    0.56
     ◆
    0.55
    ↵↵↵↵
    0.54
    </h1>
    0.54
     ">
    0.54
     アンティーク
    0.54
    </h4>
    0.53
     LEGO
    0.53
    POSITIVE LOGITS
    থিত
    0.65
    раз
    0.63
    тина
    0.61
    equivariant
    0.61
    widehat
    0.60
    usages
    0.60
    дит
    0.59
    handleChange
    0.59
    encerramento
    0.59
    刽
    0.59
    Activations Density 0.013%

    No Known Activations