© 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. 10-GEMMASCOPE-RES-131K
    7. 111694
    Prev
    Next
    INDEX
    Explanations

    introductory phrases for statements

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

    The neuron fires on short multiword, sentence-opening commentary or transition phrases (e.g. “Comparisons are made,” “No real rancher,” “The correct way to write that sentence,” “He is expected to miss,” “Cue the dreamy eyes,” etc.)

    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
     "..\..\..\
    -1.55
    JspWriter
    -1.48
     "..\..\
    -1.48
    metres
    -1.46
    ━━
    -1.46
    ཌ
    -1.45
    ￠
    -1.45
    tierten
    -1.45
    usions
    -1.43
    </caption>
    -1.41
    POSITIVE LOGITS
     that
    1.81
     or
    1.70
    0
    1.67
     tersebut
    1.45
     which
    1.33
    1
    1.27
     another
    1.24
     then
    1.23
     schlafen
    1.23
    我自己
    1.22
    Activations Density 0.093%

    No Known Activations