© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28120
    Prev
    Next
    INDEX
    Explanations

    alternatives to disabling

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

    The neuron fires strongly on main action verbs (e.g. “adopt,” “tried,” “comes,” “disable,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     whether
    -0.95
    uffff
    -0.94
     quiser
    -0.92
     Whether
    -0.92
     gelungen
    -0.90
     justice
    -0.90
     weather
    -0.89
     correos
    -0.89
     site
    -0.87
     fourth
    -0.87
    POSITIVE LOGITS
     appris
    1.23
    そちら
    1.16
    こっち
    1.07
     圖片
    1.02
    mdan
    0.99
    esthetics
    0.99
    räder
    0.96
     Tugas
    0.94
     incompar
    0.92
    alternate
    0.91
    Activations Density 0.079%

    No Known Activations