© 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. 1412
    Prev
    Next
    INDEX
    Explanations

    causing, making, thereby

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

    The neuron strongly activates on numeric tokens with decimal points (i.e. numbers/measurements).

    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
     मैंने
    0.83
     jedan
    0.78
     छह
    0.77
    했지만
    0.74
     zwölf
    0.74
     sechs
    0.73
     anunció
    0.72
     嗯
    0.70
     chaired
    0.70
     escribió
    0.70
    POSITIVE LOGITS
     ensures
    0.69
     increases
    0.67
     decreases
    0.67
     potentially
    0.62
     facilitates
    0.61
    Allows
    0.61
     further
    0.59
    从而
    0.59
     exacerb
    0.57
     minimizes
    0.57
    Activations Density 0.056%

    No Known Activations