© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129000
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric literals specifying bit‐widths in type names (e.g. the “32” in “uint32_t”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    uint type declarations

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     dejando
    -2.64
     poniendo
    -2.50
    都是
    -2.39
    ↵
    -2.30
     amplia
    -2.22
     -
    -2.20
     That
    -2.19
     แต่
    -2.16
     haciendo
    -2.14
     Roughly
    -2.09
    POSITIVE LOGITS
    塒
    2.73
     pérd
    2.67
    𓋼
    2.67
    桖
    2.67
    鱻
    2.52
     anún
    2.47
    
    2.45
     fece
    2.38
     viera
    2.38
    泐
    2.36
    Activations Density 0.003%

    No Known Activations