© 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. 66218
    Prev
    Next
    INDEX
    Explanations

    follow instruction or direction

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

    The neuron consistently activates on instances of the word “Follow” (e.g. in “Follow us on…” or “Follow this and additional works…”).

    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
    "
    -2.67
    {
    -2.52
     be
    -2.41
    ’
    -2.41
     has
    -2.33
    '
    -2.33
     se
    -2.25
     u
    -2.23
     .
    -2.19
     was
    -2.16
    POSITIVE LOGITS
    ⛦
    3.34
    谂
    3.06
     był
    2.88
    媖
    2.86
     eſſ
    2.80
    咘
    2.67
    ☫
    2.66
    إذا
    2.53
    我们
    2.52
    狁
    2.52
    Activations Density 0.044%

    No Known Activations