© 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. 10-GEMMASCOPE-RES-131K
    7. 40602
    Prev
    Next
    INDEX
    Explanations

    performing actions manually

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

    The neuron fires on occurrences of “manual” (and its variants like “manually”), i.e. words indicating a manual action or intervention.

    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
    ’
    -1.52
    bisa
    -1.38
     I
    -1.38
     نت
    -1.38
    Herb
    -1.34
    </tfoot>
    -1.32
    Filmography
    -1.31
    udere
    -1.27
    )
    -1.27
    小镇
    -1.26
    POSITIVE LOGITS
     manually
    2.02
    ”，
    1.91
    ”,
    1.72
    ——”
    1.65
    this
    1.63
     excepción
    1.62
    about
    1.55
     manual
    1.52
    ?),
    1.49
    ureusement
    1.47
    Activations Density 0.035%

    No Known Activations