© 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. 33840
    Prev
    Next
    INDEX
    Explanations

    question, explanation, crystal, reconstruction, hates

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

    This neuron responds to long alphanumeric strings (such as URL parameters or ID codes).

    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
     could
    -2.00
     should
    -1.88
     might
    -1.86
     would
    -1.74
     may
    -1.69
     as
    -1.66
     with
    -1.56
     which
    -1.52
     has
    -1.47
     for
    -1.39
    POSITIVE LOGITS
    wanted
    1.49
    quickly
    1.45
    ebly
    1.45
    wants
    1.44
    athered
    1.41
    seems
    1.41
    provides
    1.41
    mogen
    1.41
    allows
    1.40
    nade
    1.37
    Activations Density 0.005%

    No Known Activations