© 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. 22-GEMMASCOPE-RES-131K
    7. 12752
    Prev
    Next
    INDEX
    Explanations

    code syntax involving assignment and separators

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

    The neuron consistently lights up on user‐defined identifiers (i.e. variable, parameter, or function names) in code.

    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
    事です
    -1.34
    oggetto
    -1.34
     acheter
    -1.32
    rily
    -1.32
     obuv
    -1.28
     réal
    -1.24
     willen
    -1.23
     chce
    -1.22
    ホント
    -1.21
     žena
    -1.21
    POSITIVE LOGITS
     was
    1.61
     didn
    1.45
     =
    1.30
    は
    1.30
     couldn
    1.18
     should
    1.15
     still
    1.15
     is
    1.15
     had
    1.12
     will
    1.08
    Activations Density 0.006%

    No Known Activations