© 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. 57708
    Prev
    Next
    INDEX
    Explanations

    Saf followed by completions

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

    The neuron strongly activates on tokens containing the substring “saf” (e.g. SAFE, Safeway, Safranin) in any context.

    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
     Meksika
    -0.76
    SOM
    -0.76
     CALIFORNI
    -0.75
     gezond
    -0.75
    &&&
    -0.73
     memória
    -0.71
     realization
    -0.71
    tift
    -0.71
     verdaderos
    -0.70
    tats
    -0.67
    POSITIVE LOGITS
     Saf
    1.23
    Saf
    1.09
     saf
    1.05
    saf
    0.85
     SAF
    0.83
     haven
    0.79
    ).__
    0.79
    SAF
    0.78
    !("
    0.77
     wollte
    0.75
    Activations Density 0.013%

    No Known Activations