© 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. 65477
    Prev
    Next
    INDEX
    Explanations

    were included or will be

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

    The neuron strongly activates on very common English function and auxiliary words (e.g. forms of “be,” “will,” “can,” “were,” etc.), i.e. high-frequency tokens rather than specific content.

    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
     one
    -1.03
    atikan
    -1.00
     another
    -0.98
     один
    -0.94
     tett
    -0.93
    EventArgs
    -0.91
    jangan
    -0.91
    ๗
    -0.90
    ouro
    -0.90
     niin
    -0.88
    POSITIVE LOGITS
     are
    1.24
    those
    1.02
    hereinafter
    0.99
    会被
    0.92
     těch
    0.92
     eligible
    0.91
     will
    0.91
    捱
    0.90
    するもの
    0.88
     레이
    0.86
    Activations Density 0.066%

    No Known Activations