© 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. 128115
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on clause-boundary markers—especially commas and the common function/connective words that follow or precede them.

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

    despite and regardless

    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
    闈
    -1.01
    elancer
    -0.99
     Rosh
    -0.95
    stv
    -0.93
     Macbeth
    -0.88
     myös
    -0.88
    gpt
    -0.87
     fikir
    -0.86
     kalem
    -0.85
     conductores
    -0.84
    POSITIVE LOGITS
     despite
    0.89
     since
    0.88
    To
    0.82
     الرغم
    0.79
     distinct
    0.79
    　※
    0.79
    SourceChecksum
    0.79
     regard
    0.78
     läs
    0.78
     regardless
    0.77
    Activations Density 0.012%

    No Known Activations