© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 57
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on the verb “do.”

    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
     prilikom
    0.37
    Base
    0.34
               
    0.33
    OutputStream
    0.33
    ">
    0.32
    enium
    0.32
    InputStream
    0.32
    Initializer
    0.30
    "></
    0.30
    pore
    0.30
    POSITIVE LOGITS
     wonders
    0.55
     needful
    0.52
     housework
    0.50
     something
    0.48
     damage
    0.48
     justice
    0.48
    omsday
    0.47
     DAMAGE
    0.47
     nothing
    0.47
     squats
    0.47
    Activations Density 0.254%

    No Known Activations