© 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. 16004
    Prev
    Next
    INDEX
    Explanations

    hexadecimal or memory addresses

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

    The neuron spikes on uppercase hex digits (A–F) inside hexadecimal constants.

    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
     acceptez
    -1.54
    
    -1.46
     проблеми
    -1.43
     cilindros
    -1.43
     klachten
    -1.42
     illustratie
    -1.41
    aç
    -1.39
    岀
    -1.37
     farba
    -1.35
     ajustar
    -1.33
    POSITIVE LOGITS
    \
    1.77
    9
    1.59
    1
    1.59
    <i>
    1.48
    well
    1.46
    ED
    1.45
    3
    1.43
     You
    1.42
    5
    1.40
    ↵
    1.39
    Activations Density 0.002%

    No Known Activations