© 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. 64798
    Prev
    Next
    INDEX
    Explanations

    temporary files and directories

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

    The main thing this neuron does is detect references to temporary‐file or temporary‐directory constructs (e.g. “tempfile,” “mkdtemp,” “tmp,” etc.).

    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
     follows
    -1.11
     POLIC
    -0.99
    チール
    -0.98
     souhaitent
    -0.96
    の部分
    -0.96
     desarrolló
    -0.95
     otorga
    -0.93
     volet
    -0.93
     prénom
    -0.92
    以上に
    -0.92
    POSITIVE LOGITS
     Temporary
    1.54
     temporary
    1.52
    Temporary
    1.50
    temporary
    1.37
     tempfile
    1.35
     createState
    1.23
     {
    1.20
    ("__
    1.11
     tempor
    1.08
     ephemeral
    1.08
    Activations Density 0.023%

    No Known Activations