© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 48889
    Prev
    Next
    INDEX
    Explanations

    platform data

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

    The neuron fires strongly on underscore characters—i.e. snake_case identifiers or tokens containing “_” in code.

    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
     habit
    -0.82
     vorbere
    -0.79
     valentine
    -0.77
    Cycles
    -0.77
     uſ
    -0.77
    Judas
    -0.76
     Habit
    -0.75
     Mula
    -0.75
    袈
    -0.74
    🌆
    -0.74
    POSITIVE LOGITS
     platform
    1.00
     afraid
    0.87
     feared
    0.85
    imentos
    0.84
     platforms
    0.83
    prés
    0.82
     Platform
    0.81
    Plat
    0.77
     she
    0.77
    គ
    0.76
    Activations Density 0.013%

    No Known Activations