© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 42363
    Prev
    Next
    INDEX
    Explanations

    programming data types and syntax

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

    The neuron activates on code tokens that name or invoke types—especially type casts or generic/type‐parameter constructions (e.g. “as String”, “as UI.ViewController”, “new T()”).

    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
     to
    -1.63
     you
    -1.50
     as
    -1.34
     and
    -1.30
     memutus
    -1.28
     in
    -1.27
    </b>
    -1.27
     of
    -1.20
     мес
    -1.19
    入った
    -1.16
    POSITIVE LOGITS
     όταν
    1.73
    This
    1.49
    quando
    1.48
     limitaciones
    1.45
     pérdidas
    1.45
    因為
    1.41
     ghiaccio
    1.40
     scuro
    1.38
    requently
    1.37
     this
    1.36
    Activations Density 0.009%

    No Known Activations