© 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. 108071
    Prev
    Next
    INDEX
    Explanations

    administrate, penetrate, registries

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

    The neuron strongly activates on the subword “penet,” i.e. occurrences of that root in words like penetrate, penetration, or penetrating.

    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
    ZE
    -0.86
    resizable
    -0.85
    ölcs
    -0.75
     secure
    -0.75
     tala
    -0.72
    madeus
    -0.71
    Delimiter
    -0.71
    authenticate
    -0.71
    secure
    -0.70
    锆
    -0.69
    POSITIVE LOGITS
    tration
    1.27
    ration
    1.21
    trate
    1.05
    rated
    1.03
    ation
    1.00
    ATION
    0.94
    rating
    0.92
    rations
    0.90
    ations
    0.85
    ratings
    0.84
    Activations Density 0.052%

    No Known Activations