© 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. 22-GEMMASCOPE-RES-131K
    4. 35804
    Prev
    Next
    INDEX
    Explanations

    code structure elements

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

    This neuron detects code tokens that name types or interfaces—i.e. capitalized identifiers like class, interface, or model names.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     niega
    -1.73
     akurat
    -1.70
     récolte
    -1.59
     sostu
    -1.59
     rêver
    -1.57
     relâche
    -1.56
     kuiten
    -1.56
     advierte
    -1.55
     tomaron
    -1.53
     célè
    -1.47
    POSITIVE LOGITS
    <eos>
    1.84
     &
    1.38
     would
    1.35
     one
    1.26
    ire
    1.25
     es
    1.25
    ien
    1.23
    ins
    1.23
    ub
    1.23
    ur
    1.23
    Activations Density 0.213%

    No Known Activations