© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 15141
    Prev
    Next
    INDEX
    Explanations

    after a disastrous or determined event

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

    The neuron strongly activates on tokens within extended, detailed instruction or how-to sections—that is, it flags long, procedural technical directions.

    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
     defines
    0.82
     distinguishing
    0.79
    awang
    0.74
    ول
    0.73
     defining
    0.73
     Defines
    0.73
    ع
    0.72
     enlarging
    0.70
     directs
    0.69
     defined
    0.68
    POSITIVE LOGITS
    tank
    0.86
    gmail
    0.82
     sandalwood
    0.81
    tracker
    0.79
    turtle
    0.78
    tan
    0.77
    thrift
    0.77
    𝚐
    0.77
    trä
    0.76
    train
    0.75
    Activations Density 0.000%

    No Known Activations