© 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. 10-GEMMASCOPE-RES-131K
    7. 44915
    Prev
    Next
    INDEX
    Explanations

    .h or .hpp files

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

    The neuron is highly sensitive to file‐include directives—specifically the header‐file names (and their extensions) used in #include lines.

    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
     In
    -1.50
     expedite
    -1.38
     to
    -1.38
     varias
    -1.33
     At
    -1.29
     eleven
    -1.28
     with
    -1.27
     That
    -1.24
     Introduction
    -1.23
    чему
    -1.23
    POSITIVE LOGITS
     yaka
    1.38
     hunne
    1.36
     courir
    1.31
     sociala
    1.30
     deserto
    1.30
     noft
    1.30
     kapture
    1.30
     haer
    1.29
     pering
    1.29
     Azi
    1.28
    Activations Density 0.051%

    No Known Activations