© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14681
    Prev
    Next
    INDEX
    Explanations

    using a tool or method

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

    The neuron is detecting procedural‐method markers—words like “using,” “with,” “by,” etc., that introduce or describe steps of a method or process.

    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
    ichord
    -1.06
     gesamte
    -1.04
    參考文獻
    -1.02
     sämtliche
    -1.01
     entire
    -1.00
     demás
    -0.99
     모든
    -0.98
     Familienname
    -0.98
    quam
    -0.96
    贷
    -0.96
    POSITIVE LOGITS
     by
    1.57
     a
    1.34
     an
    1.31
     using
    1.30
     yine
    1.29
     another
    1.15
     either
    1.14
     помощью
    1.12
     cuidad
    1.09
     vence
    1.09
    Activations Density 0.152%

    No Known Activations