© 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. 10-GEMMASCOPE-RES-131K
    7. 128171
    Prev
    Next
    INDEX
    Explanations

    toll on, echoes

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

    The main thing this neuron does is detect mentions of something “taking a toll” or causing a harmful impact.

    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
    しよう
    -1.62
     Sometimes
    -1.57
     Are
    -1.52
     Many
    -1.45
    YECTO
    -1.39
     Damit
    -1.37
     Why
    -1.36
    憮
    -1.34
    綑
    -1.34
     Contrary
    -1.34
    POSITIVE LOGITS
     alkoholu
    1.61
    乘客
    1.45
    なども
    1.45
     también
    1.45
    legte
    1.42
    también
    1.42
    نور
    1.41
    方は
    1.40
     konkuren
    1.38
     renou
    1.38
    Activations Density 0.014%

    No Known Activations