© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 38618
    Prev
    Next
    INDEX
    Explanations

    file formats or metadata

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

    The neuron is keyed to any token containing digits (i.e. it fires on numeric literals/version numbers).

    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
     Landis
    -0.69
     Frankie
    -0.69
    aronder
    -0.66
    zzar
    -0.65
     rồng
    -0.64
    alat
    -0.63
    jboss
    -0.63
    爺
    -0.60
    WESTERN
    -0.60
    在中国
    -0.60
    POSITIVE LOGITS
    cratic
    0.69
    obiety
    0.67
     ചെയ
    0.66
    敷
    0.66
    pads
    0.65
     verlassen
    0.64
     базе
    0.63
     dicht
    0.63
    mentaire
    0.63
    frees
    0.62
    Activations Density 0.075%

    No Known Activations