© 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. 22-GEMMASCOPE-RES-131K
    7. 73483
    Prev
    Next
    INDEX
    Explanations

    what and where questions

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

    The neuron fires strongly on the first words of sentences or clauses (i.e. discourse‐marker tokens at sentence or clause beginnings).

    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
     zentral
    -0.90
    tiness
    -0.86
     srdce
    -0.86
     anti
    -0.84
     niebla
    -0.83
    anti
    -0.83
     pierna
    -0.81
    oczes
    -0.80
    abou
    -0.80
     wall
    -0.80
    POSITIVE LOGITS
    ombra
    0.94
     thereunder
    0.92
     Congreg
    0.91
     Caledonia
    0.91
     abstin
    0.90
    টি
    0.90
     mwaka
    0.90
     kyse
    0.90
     portables
    0.89
     duos
    0.89
    Activations Density 0.184%

    No Known Activations