© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 32600
    Prev
    Next
    INDEX
    Explanations

    programming constructs and keywords

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

    The neuron is primarily detecting numeric tokens and other number‐like expressions (e.g. counts, dates, years, monetary amounts).

    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
    NOWLEDGE
    -0.78
    kým
    -0.73
     fote
    -0.73
     Sándor
    -0.72
     Rans
    -0.71
    шной
    -0.71
     Several
    -0.70
     ecles
    -0.70
     Ayam
    -0.69
     await
    -0.68
    POSITIVE LOGITS
     cały
    0.92
    arya
    0.90
    ונים
    0.89
    Cyber
    0.89
    踉
    0.88
     Cyber
    0.84
     lebte
    0.83
     Potomac
    0.82
    аа
    0.82
    Kontrola
    0.81
    Activations Density 0.004%

    No Known Activations