© 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. 26738
    Prev
    Next
    INDEX
    Explanations

    dynamic memory allocation

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

    The neuron is keyed to C++ dynamic‐allocation sites, specifically spotting occurrences of the “new” operator.

    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
    ↵↵
    -1.36
     différentes
    -1.30
     apareceu
    -1.22
     recebeu
    -1.19
    ared
    -1.19
    ようで
    -1.15
     definição
    -1.15
    Ꭳ
    -1.14
     criação
    -1.13
    PyExc
    -1.13
    POSITIVE LOGITS
     at
    1.41
     during
    1.34
    hängigkeit
    1.23
    ＝＝
    1.23
     wisatawan
    1.21
    目的地
    1.19
     amidst
    1.19
     organizar
    1.18
     there
    1.17
     now
    1.16
    Activations Density 0.003%

    No Known Activations