© 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. 15754
    Prev
    Next
    INDEX
    Explanations

    code and technical terms

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

    The neuron tends to spike on isolated, content-heavy tokens—especially rare, domain-specific terms (e.g. specialized technical or scientific words).

    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
     unless
    -0.92
     плюс
    -0.79
     or
    -0.79
     Tern
    -0.77
     mV
    -0.75
    unless
    -0.74
    jia
    -0.74
     gg
    -0.72
     nazionali
    -0.70
     puls
    -0.69
    POSITIVE LOGITS
    StructEnd
    1.05
     terle
    0.90
     unido
    0.88
     péro
    0.85
     recomendado
    0.84
    ologue
    0.84
    Normals
    0.84
     rozwój
    0.83
    总体
    0.83
     isFirst
    0.83
    Activations Density 0.060%

    No Known Activations