© 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. 34-GEMMASCOPE-RES-131K
    7. 34538
    Prev
    Next
    INDEX
    Explanations

    Hex bytes and numbers

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

    The neuron is detecting numeric tokens—individual digits or multi‐digit literals (e.g. array sizes, numeric constants, numbers in `#define`s).

    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
     absorbed
    -0.77
     dzień
    -0.76
     aynı
    -0.75
    цька
    -0.70
     czarna
    -0.70
    эг
    -0.69
     modos
    -0.69
    rotz
    -0.69
    bf
    -0.69
    BigInteger
    -0.69
    POSITIVE LOGITS
    ourites
    0.95
    országban
    0.80
    BlockSize
    0.79
    èvre
    0.77
    NamedQueries
    0.77
    UTES
    0.77
     COMMAND
    0.76
     grapes
    0.75
     sizeof
    0.75
    sizeof
    0.72
    Activations Density 0.018%

    No Known Activations