© 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. 90939
    Prev
    Next
    INDEX
    Explanations

    from the

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

    The neuron strongly activates on numeric tokens—especially multi‐digit numbers like years, version numbers, or case/document IDs.

    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
     βά
    -0.82
     posibilidad
    -0.77
     الأول
    -0.74
    Sirius
    -0.74
    cieli
    -0.73
    registrer
    -0.73
     dritte
    -0.72
     EXTER
    -0.71
    utilisons
    -0.71
    czegó
    -0.70
    POSITIVE LOGITS
    ناف
    0.73
    weightedMode
    0.71
     sorgt
    0.68
     solutions
    0.66
    lanted
    0.65
    alumnos
    0.65
    빼
    0.64
    VAC
    0.63
     USE
    0.63
    イト
    0.63
    Activations Density 0.065%

    No Known Activations