© 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. 1464
    Prev
    Next
    INDEX
    Explanations

    abstract concepts

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

    The neuron strongly activates on long, abstract, Latin-derived nouns (and other weighty conceptual words) conveying philosophical, emotional, or metaphysical ideas.

    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
    0
    0.71
     cheeky
    0.66
     whopping
    0.63
     அவரின்
    0.63
     العديد
    0.62
     freestyle
    0.61
     juicy
    0.61
     sneaky
    0.61
     Xbox
    0.60
     perks
    0.60
    POSITIVE LOGITS
     siglos
    0.60
    ,/
    0.59
     whence
    0.59
    urrection
    0.59
     Aś
    0.59
     ಹೃ
    0.58
     ritor
    0.57
     aussitôt
    0.56
     poeta
    0.55
    ごと
    0.55
    Activations Density 0.001%

    No Known Activations