© 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. 34-GEMMASCOPE-RES-131K
    7. 4237
    Prev
    Next
    INDEX
    Explanations

    numbers and specific words

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

    The neuron primarily detects numeric tokens (digits and multi‐digit number sequences such as years, case numbers, and section citations).

    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
    Vigil
    -0.84
    værk
    -0.81
    reathe
    -0.80
    Svara
    -0.77
     Vigil
    -0.75
     avoit
    -0.74
    נְ
    -0.73
    Nevertheless
    -0.71
     nimmer
    -0.69
    🪳
    -0.69
    POSITIVE LOGITS
    LEA
    0.77
     Felt
    0.75
     feel
    0.73
    abbix
    0.73
    felt
    0.72
    界的
    0.72
     пру
    0.71
    leads
    0.71
     behold
    0.71
     relates
    0.70
    Activations Density 0.143%

    No Known Activations