© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3943
    Prev
    Next
    INDEX
    Explanations

    state of the art

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

    The neuron strongly activates on uncommon, domain-specific proper names or jargon terms (e.g. “spectrum,” “Idolverse,” “canon,” “blogiosphere,” “GAS”).

    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
     carinatis
    0.60
    નના
    0.57
    㴌
    0.56
    την
    0.55
    Перед
    0.53
     पुष्पा
    0.52
    䜣
    0.51
    После
    0.51
    achet
    0.51
    मीन
    0.50
    POSITIVE LOGITS
    /
    0.88
    (?)
    0.84
     (
    0.82
     (?)
    0.82
    /"
    0.77
     isn
    0.77
     tbh
    0.74
     I
    0.72
     wouldn
    0.72
     /
    0.72
    Activations Density 0.001%

    No Known Activations