© 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. 22-GEMMASCOPE-RES-131K
    7. 129988
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on capitalized type‐identifier tokens in code—especially user‐defined names in declarations (e.g. contract names, struct names, and similar identifiers).

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

    code comments and structures

    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
    FirebaseUser
    -1.04
    这些
    -1.02
     either
    -1.01
    either
    -0.99
     might
    -0.99
     returned
    -0.95
    $(\
    -0.92
    由于
    -0.90
    -//
    -0.90
    Если
    -0.90
    POSITIVE LOGITS
    Ery
    1.20
     ingrediente
    1.19
    Karol
    1.11
    眼泪
    1.09
    ñó
    1.09
     tranche
    1.09
    violin
    1.05
    後で
    1.04
    ודה
    1.02
     meningkatkan
    1.02
    Activations Density 0.002%

    No Known Activations