© 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. 75424
    Prev
    Next
    INDEX
    Explanations

    worker in

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

    The neuron fires on tokens that are purely numeric (or largely numeric) sequences, i.e. numbers, line‐number references, years, zip/postal codes, and other digit strings.

    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
    Źródło
    -1.02
     Those
    -0.91
     Although
    -0.90
     Almost
    -0.89
     first
    -0.86
     Nossa
    -0.85
     exquisitely
    -0.84
     gols
    -0.83
     had
    -0.82
     Because
    -0.82
    POSITIVE LOGITS
     electricidad
    0.97
     Determines
    0.90
     necessidade
    0.90
     ausreichend
    0.90
     mw
    0.87
     behandelt
    0.87
     |-
    0.86
     bisogno
    0.84
     atendimento
    0.84
     necesarios
    0.83
    Activations Density 0.005%

    No Known Activations