© 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. 108134
    Prev
    Next
    INDEX
    Explanations

    link span

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

    The neuron strongly activates on numeric tokens—scores, counts, percentages, or other multi‐digit 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
     his
    -1.16
     himself
    -0.88
     beliau
    -0.84
     personales
    -0.84
    ikuti
    -0.83
    ":[{
    -0.82
    šel
    -0.81
    سجيلات
    -0.80
     он
    -0.79
    entlichen
    -0.79
    POSITIVE LOGITS
     their
    2.02
     its
    1.50
     collectively
    1.27
     collective
    1.20
     compared
    1.19
     Their
    1.15
     deres
    1.13
     they
    1.13
     themselves
    1.12
     kanilang
    1.12
    Activations Density 0.010%

    No Known Activations