© 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. 705
    Prev
    Next
    INDEX
    Explanations

    Q: or /*

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

    The neuron strongly activates on long runs of hexadecimal digit sequences (e.g. hex dumps) in the 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
     during
    -0.87
    During
    -0.84
     During
    -0.79
     Notably
    -0.78
     However
    -0.77
     nostri
    -0.77
     Initially
    -0.77
     seemingly
    -0.75
     or
    -0.75
     самая
    -0.73
    POSITIVE LOGITS
    optimized
    0.93
     combin
    0.92
    ítás
    0.91
    ीक
    0.91
     Boucher
    0.90
    ConstraintLayout
    0.90
    USET
    0.87
    IGNORE
    0.84
    compare
    0.83
    tinyos
    0.83
    Activations Density 0.005%

    No Known Activations