© 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. 10-GEMMASCOPE-RES-131K
    7. 8612
    Prev
    Next
    INDEX
    Explanations

    place names and phrases

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

    The neuron detects sequences of hexadecimal digits (as found in Unicode escape or similar hex‐encoded 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_10/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
     and
    -1.95
     cucchiai
    -1.84
    鈮
    -1.72
     adresu
    -1.70
    kenstock
    -1.70
     autory
    -1.63
     to
    -1.60
    şam
    -1.59
     podoba
    -1.59
     zvuk
    -1.52
    POSITIVE LOGITS
     also
    2.09
     of
    1.73
    This
    1.67
    It
    1.66
     ähn
    1.65
     other
    1.65
    Also
    1.60
    但也
    1.58
    ens
    1.58
    o
    1.51
    Activations Density 0.042%

    No Known Activations