© 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. 34-GEMMASCOPE-RES-131K
    7. 30514
    Prev
    Next
    INDEX
    Explanations

    citations and questions

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

    The neuron is picking out code‐style identifiers and technical tokens (e.g. class and function names, CamelCase terms, library/component names) in source or header text.

    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
     takeout
    -0.75
     fracturing
    -0.73
     hurrying
    -0.73
    omie
    -0.70
     Argy
    -0.68
     polygonal
    -0.68
     बाह
    -0.67
     Orche
    -0.66
    itete
    -0.65
    odha
    -0.65
    POSITIVE LOGITS
     buses
    0.83
    tokio
    0.83
    дере
    0.81
    endpoint
    0.79
    brium
    0.79
     ks
    0.76
    -|
    0.75
     worker
    0.75
     topology
    0.75
    topology
    0.74
    Activations Density 0.033%

    No Known Activations