© 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. 33264
    Prev
    Next
    INDEX
    Explanations

    code comments, questions, or transmissions

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

    The neuron strongly activates on non‐lexical reference tokens—things like standalone numbers, uppercase acronyms or codes, and punctuation markers (e.g. semicolons).

    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
     here
    -0.93
     for
    -0.92
     sintomi
    -0.90
     of
    -0.78
    这里
    -0.77
    劭
    -0.75
     làng
    -0.75
     CF
    -0.74
    PageModule
    -0.73
     having
    -0.73
    POSITIVE LOGITS
    τές
    0.87
    Moo
    0.86
     vỡ
    0.85
    yden
    0.83
    irce
    0.82
    Eugen
    0.81
    IActionResult
    0.80
    Lith
    0.79
    ,:);
    0.79
    лик
    0.79
    Activations Density 0.000%

    No Known Activations