© 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. 7733
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on interrogative “Which” questions (and their nearby content words) in technical or Q&A–style sentences.

    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
     หา
    0.89
     Huan
    0.88
     NADPH
    0.81
    𝗘
    0.79
     항상
    0.78
     모르
    0.78
     이유는
    0.78
     Ե
    0.77
     insanın
    0.77
     ϒ
    0.76
    POSITIVE LOGITS
     welke
    0.81
                            
    0.79
    ov
    0.73
    "',
    0.73
     turns
    0.71
     centerpiece
    0.69
     harbors
    0.69
     laborer
    0.69
    iv
    0.68
    这将
    0.67
    Activations Density 0.015%

    No Known Activations