© 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. 129508
    Prev
    Next
    INDEX
    Explanations

    The neuron is highly sensitive to the symbol “μ” (mu), i.e. references to the Greek-letter parameter μ (often a mean) in formulas or parameter names.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    mu prefix

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    .”
    -2.50
    具有
    -2.47
    x
    -2.39
     directivos
    -2.25
    需要
    -2.25
    有效的
    -2.19
     اذا
    -2.19
    大部分
    -2.16
    });
    -2.14
     verbess
    -2.14
    POSITIVE LOGITS
    *
    2.81
    骘
    2.70
    鹪
    2.63
    ೣ
    2.59
    虓
    2.47
     "
    2.45
     الأولى
    2.44
     脚
    2.44
    \
    2.39
    ributes
    2.33
    Activations Density 0.032%

    No Known Activations