© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66625
    Prev
    Next
    INDEX
    Explanations

    therefore

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

    The neuron detects words that point ahead to examples or supplementary material (e.g. “examples,” “shown,” “below,” “discussed,” “snippet,” “explain”).

    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
    js
    -0.81
    认为
    -0.77
     хозяйства
    -0.76
    อะไร
    -0.75
     Š
    -0.75
    stí
    -0.74
    認為
    -0.73
    zaki
    -0.73
     נפ
    -0.73
     qualquer
    -0.73
    POSITIVE LOGITS
     them
    1.21
     accordingly
    1.20
    はその
    1.13
     därför
    1.07
     ấy
    1.07
     derfor
    1.06
     therefore
    1.04
     therefor
    1.04
     hivyo
    1.02
     galite
    1.01
    Activations Density 0.084%

    No Known Activations