© 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. 33324
    Prev
    Next
    INDEX
    Explanations

    resetting or reinstalling software

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

    The neuron fires strongly on occurrences of the word “reset” (especially in technical or instructional contexts).

    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
    </h1>
    -1.22
    是一種
    -0.99
     unveils
    -0.98
    𝚖
    -0.96
    </strong>
    -0.95
    wid
    -0.93
    E
    -0.93
     aq
    -0.93
     ネイビー
    -0.92
    还需要
    -0.92
    POSITIVE LOGITS
     reset
    1.70
     Reset
    1.36
     resetting
    1.22
     tulis
    1.14
    RESET
    1.14
    haviours
    1.13
     fresh
    1.13
     resets
    1.11
    reset
    1.10
     clean
    1.08
    Activations Density 0.034%

    No Known Activations