© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 45470
    Prev
    Next
    INDEX
    Explanations

    thing to do

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

    This neuron detects phrases of the form “… thing to do” (e.g. “the right thing to do,” “best thing to do,” “one thing to do,” “nothing to do”).

    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
    ╳
    -0.81
     Colleen
    -0.79
     Temporada
    -0.76
     Deploy
    -0.75
     sydney
    -0.74
    ¥
    -0.73
     Donate
    -0.73
    ıs
    -0.73
     funkcji
    -0.72
    _______
    -0.72
    POSITIVE LOGITS
     doing
    2.63
    Doing
    1.69
     do
    1.69
     делать
    1.59
    doing
    1.57
     todo
    1.47
     Doing
    1.46
    要做
    1.46
     לעשות
    1.45
     doin
    1.40
    Activations Density 0.079%

    No Known Activations