© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 102389
    Prev
    Next
    INDEX
    Explanations

    `shields`

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

    This neuron activates on unusually long tokens—especially long programming identifiers or multi-character names (e.g. PascalCase/CamelCase names) 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
    ynnwys
    -0.72
    धिक
    -0.71
    类型
    -0.70
    Rab
    -0.69
     Ор
    -0.67
     adopted
    -0.67
    河
    -0.66
    ダウンロード
    -0.65
    żym
    -0.65
     완
    -0.65
    POSITIVE LOGITS
    greens
    0.69
     Deuter
    0.69
     erlebt
    0.69
    薛
    0.66
    edar
    0.66
    agens
    0.65
     measles
    0.65
    Seeker
    0.64
     piecewise
    0.63
     Decke
    0.63
    Activations Density 0.048%

    No Known Activations