© 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. 3321
    Prev
    Next
    INDEX
    Explanations

    when/where needed or possible

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

    The neuron is specialized to detect numeric references and statute‐style citations (numbers, percentages, thresholds, or legal section references).

    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
     may
    0.77
     will
    0.73
     czyli
    0.67
     hvad
    0.67
     dapat
    0.66
     lahko
    0.66
     derfor
    0.66
    을
    0.66
     WILL
    0.65
     შეგიძლიათ
    0.65
    POSITIVE LOGITS
    soever
    0.66
    Compared
    0.66
    accompanied
    0.66
    eel
    0.65
    compared
    0.63
     условии
    0.61
    ともに
    0.60
    Política
    0.59
    concerned
    0.59
    ্বেও
    0.58
    Activations Density 0.079%

    No Known Activations