© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128077
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric tokens (integer or floating-point literals) in the code or text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    code variables and identifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     góry
    -1.11
     sostuvo
    -1.05
     auswählen
    -0.91
     peintures
    -0.90
    ığınız
    -0.88
     lâng
    -0.87
     eigenes
    -0.82
     Ofrece
    -0.81
    過ご
    -0.81
     unbedingt
    -0.81
    POSITIVE LOGITS
    vancing
    0.99
    vances
    0.96
     exacta
    0.93
    advanced
    0.87
     judicia
    0.86
     Unidas
    0.86
    راحل
    0.86
    onDelete
    0.86
    ولوج
    0.86
    र्ड
    0.85
    Activations Density 0.015%

    No Known Activations