© 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. 34-GEMMASCOPE-RES-131K
    7. 117535
    Prev
    Next
    INDEX
    Explanations

    ultra marathon runner

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

    The neuron is selectively activating on words from the main body of a blog post (the narrative text) and suppressing tokens from headers, dates, menus, and other boilerplate.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     lmfao
    -0.76
     LMAO
    -0.74
    Create
    -0.70
    講師
    -0.69
    އ
    -0.69
     Fische
    -0.68
     😭😭
    -0.68
     корпу
    -0.67
     PPI
    -0.66
     savings
    -0.65
    POSITIVE LOGITS
     ultra
    1.56
    Ultra
    1.41
     Ultra
    1.37
     ultras
    1.34
    ultra
    1.30
     ul
    1.28
     ультра
    1.23
     runnable
    1.16
    ウルトラ
    1.11
     ultrason
    0.98
    Activations Density 0.020%

    No Known Activations