© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 42492
    Prev
    Next
    INDEX
    Explanations

    currency trading

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

    The neuron is firing on small character‐sequence fragments (e.g. “rit,” “ap,” “cr,” etc.), i.e. it’s detecting short character n-grams inside words.

    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
     our
    -1.75
     what
    -1.53
     who
    -1.53
     several
    -1.52
     some
    -1.51
     how
    -1.50
     make
    -1.49
     do
    -1.49
     on
    -1.48
     every
    -1.46
    POSITIVE LOGITS
    лтемелер
    1.85
    ——”
    1.50
     ナイロン
    1.48
     sonde
    1.44
    GRAM
    1.44
     pè
    1.44
     gratis
    1.42
     henne
    1.41
    "",
    1.41
     bede
    1.38
    Activations Density 0.115%

    No Known Activations