© 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. 13127
    Prev
    Next
    INDEX
    Explanations

    payment and training

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

    The neuron is keyed to words naming or pointing to the participants or agents in an interaction—e.g. “user,” “you,” “trainer,” “bilaterally,” 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_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
     we
    -1.38
     our
    -0.90
    繋がり
    -0.89
     הָ
    -0.88
    zungen
    -0.83
    違います
    -0.81
    ⎪
    -0.81
    のち
    -0.80
    (''
    -0.78
    벅
    -0.78
    POSITIVE LOGITS
    可以
    1.01
     nettoyer
    0.99
     Terbaik
    0.98
     employers
    0.96
     jestem
    0.94
    通常
    0.94
     powinien
    0.94
    Angeb
    0.93
    ıyorum
    0.92
    Modell
    0.91
    Activations Density 0.042%

    No Known Activations