© 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. 6599
    Prev
    Next
    INDEX
    Explanations

    abstract concepts and qualities

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

    The neuron fires most strongly on abstract, thematic nouns that denote key concepts or subjects in a discourse.

    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
     அவரின்
    0.80
    뤘
    0.75
     bactéries
    0.73
    ICAGO
    0.72
     می‌کن
    0.72
     تیل
    0.70
     avocados
    0.70
    췄
    0.70
    rijven
    0.69
     moniker
    0.68
    POSITIVE LOGITS
    which
    0.96
     which
    0.91
    しかも
    0.74
     quale
    0.71
    ซึ่ง
    0.70
    Which
    0.69
     Which
    0.67
     indispensable
    0.67
     wich
    0.66
     ซึ่ง
    0.64
    Activations Density 0.155%

    No Known Activations