© 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. 764
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on tokens describing video‐game play and quest activity (e.g. entering an act, joining games, completing quests, walking to levels/temples, etc.).

    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
    ǜ
    0.61
    accam
    0.54
    olverine
    0.52
    impulan
    0.50
    ിലും
    0.49
    ücksicht
    0.49
    indari
    0.48
     ഹോളി
    0.48
    გუ
    0.48
    ueve
    0.48
    POSITIVE LOGITS
    ↵
    0.62
     the
    0.60
     .
    0.58
     The
    0.55
                    
    0.55
     that
    0.55
     "
    0.54
            
    0.53
           
    0.52
    the
    0.51
    Activations Density 0.393%

    No Known Activations