© 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. 5813
    Prev
    Next
    INDEX
    Explanations

    performance, stability, limits, gradients

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

    Explanation of neuron 4 behavior: the main thing this neuron does is find words that denote technical specifications or measurable performance metrics (e.g., range, consistency, accuracy, quality).

    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
    elit
    0.69
     Только
    0.68
     পৃথিবী
    0.67
    camouflage
    0.66
     вскоре
    0.66
    Эта
    0.66
    aprendizaje
    0.66
    substantial
    0.66
    threatening
    0.64
     здравоохранения
    0.64
    POSITIVE LOGITS
     kwenye
    0.95
     across
    0.95
     Across
    0.88
     beim
    0.87
     هنگام
    0.86
     lors
    0.84
     during
    0.80
     trên
    0.79
     when
    0.77
    ในการ
    0.77
    Activations Density 0.305%

    No Known Activations