© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 127587
    Prev
    Next
    INDEX
    Explanations

    linked list and tree structure terms

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

    The neuron fires on occurrences of common pointer‐field identifiers in data‐structure code (e.g. “head,” “tail,” “root”).

    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
    </u>
    -2.14
     to
    -2.05
     '
    -1.76
    情形
    -1.58
     ').
    -1.55
    “
    -1.54
     permite
    -1.52
     them
    -1.40
     σ
    -1.36
    OOC
    -1.36
    POSITIVE LOGITS
    Отлич
    1.59
    Ещё
    1.57
    Опыт
    1.53
    Друзья
    1.46
    Анти
    1.45
    Хороший
    1.45
    Противо
    1.44
    Иск
    1.42
     theat
    1.41
    Хорошо
    1.41
    Activations Density 0.004%

    No Known Activations