© 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. 34-GEMMASCOPE-RES-131K
    7. 108269
    Prev
    Next
    INDEX
    Explanations

    over sampling, overwriting, finding out

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

    The neuron strongly activates on formal technical or academic terminology (e.g. mathematical/statistical paper language), effectively flagging scientific jargon.

    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
     viata
    -0.79
     papir
    -0.72
    Pump
    -0.72
    нымъ
    -0.71
    queado
    -0.71
    weighed
    -0.71
     teis
    -0.68
     matemática
    -0.68
     CLASSIC
    -0.67
    prove
    -0.66
    POSITIVE LOGITS
     learn
    0.89
     fits
    0.81
     fitted
    0.81
     wächst
    0.80
    learn
    0.77
     grow
    0.76
     Lern
    0.76
     Erd
    0.76
     refreshing
    0.76
     dành
    0.74
    Activations Density 0.033%

    No Known Activations