© 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. 31-GEMMASCOPE-2-RES-16K
    7. 9075
    Prev
    Next
    INDEX
    Explanations

    configuration settings

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

    The neuron fires most strongly on longer abstract or technical nouns ending in suffixes like “-ion,” “-ity,” or “-ence” (e.g., configuration, confidentiality, conference).

    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
    肯
    0.74
     covari
    0.66
     throttle
    0.63
    estyle
    0.60
     सुझा
    0.59
    pom
    0.58
    ite
    0.58
    ćem
    0.58
     Thro
    0.56
    一方で
    0.56
    POSITIVE LOGITS
    Conf
    0.78
    urados
    0.76
     conf
    0.72
     Conf
    0.70
    雀
    0.70
    édération
    0.65
    urazione
    0.64
     confess
    0.63
    licts
    0.63
    conf
    0.62
    Activations Density 0.072%

    No Known Activations