© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 79111
    Prev
    Next
    INDEX
    Explanations

    names followed by actions

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

    The neuron strongly activates on very common, high-frequency tokens—especially short function words and pronouns (e.g. “he,” “to,” “had,” “an,” “off,” “pass”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     between
    -0.95
    kc
    -0.91
     podían
    -0.89
    питан
    -0.88
     desglose
    -0.87
     frühen
    -0.85
     that
    -0.85
     menjadi
    -0.84
    开口道
    -0.82
     témo
    -0.82
    POSITIVE LOGITS
     executes
    1.00
     cleanly
    0.93
    ಿದೆ
    0.93
     bair
    0.92
     принять
    0.92
    ஂ
    0.88
     uppercase
    0.88
     venter
    0.86
    ἣ
    0.85
     inexp
    0.85
    Activations Density 0.017%

    No Known Activations