© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 128611
    Prev
    Next
    INDEX
    Explanations

    nonvolatile memory access

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

    The neuron fires on mentions of non-volatile semiconductor memory technologies (e.g. terms like “non-volatile,” “EEPROM,” “flash memory,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    擄
    -1.48
    举动
    -1.46
     things
    -1.46
    就會
    -1.30
    Pee
    -1.30
    回應
    -1.29
    徬
    -1.28
    ごちそうさまでした
    -1.27
     хотел
    -1.26
    態度
    -1.22
    POSITIVE LOGITS
     réun
    1.27
    梭
    1.20
    坑
    1.20
    Гц
    1.20
     désactiv
    1.20
     endommag
    1.19
     Chang
    1.16
     vistazo
    1.14
    ensores
    1.14
    <bos>
    1.13
    Activations Density 0.070%

    No Known Activations