© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 65662
    Prev
    Next
    INDEX
    Explanations

    code symbols and common characters

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

    The neuron strongly activates on individual tokens that include digits (i.e. numbers, numeric constants, hex values, bracketed counts, etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     of
    -1.36
     évé
    -1.34
     navigating
    -1.28
    
    -1.27
    .
    -1.25
     hörte
    -1.23
     aéreo
    -1.20
     attracting
    -1.20
     asegurarse
    -1.18
    ۠
    -1.18
    POSITIVE LOGITS
    w
    1.69
    th
    1.54
    r
    1.51
    i
    1.44
    in
    1.38
    ");
    1.37
    y
    1.34
     }
    1.33
    }{
    1.29
    h
    1.28
    Activations Density 0.295%

    No Known Activations