© 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. 22-GEMMASCOPE-RES-131K
    7. 39099
    Prev
    Next
    INDEX
    Explanations

    conjunctions like 'and' or 'or'

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

    The neuron strongly activates on numeric tokens (quantitative values and measurements).

    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
     instead
    -0.87
     etc
    -0.81
     ziren
    -0.79
     干
    -0.79
     样
    -0.77
     قابلیت
    -0.76
    ⅽ
    -0.76
     Arhi
    -0.76
     amongst
    -0.75
    ède
    -0.75
    POSITIVE LOGITS
     assise
    0.94
     גם
    0.94
    estina
    0.90
    何が
    0.88
    RTLD
    0.86
    にほんブログ村
    0.84
    aiso
    0.83
     apparaître
    0.82
     også
    0.82
     camisas
    0.82
    Activations Density 0.137%

    No Known Activations