© 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. 34-GEMMASCOPE-RES-131K
    7. 99565
    Prev
    Next
    INDEX
    Explanations

    domain of power and control

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

    This neuron detects prepositional phrases that specify scope or limits—e.g. “within the…,” “in the interest of…,” or “outside the….”

    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
    romptu
    -0.86
    outa
    -0.84
    gezogen
    -0.83
    Jaco
    -0.81
    thig
    -0.79
    ROP
    -0.79
    theilen
    -0.78
    𝗪
    -0.77
     Blaise
    -0.77
    ſſen
    -0.76
    POSITIVE LOGITS
     power
    1.79
     powers
    1.50
     within
    1.32
     capabilities
    1.27
     capacity
    1.22
     abilities
    1.20
     ability
    1.18
    power
    1.18
     capability
    1.16
     purview
    1.16
    Activations Density 0.016%

    No Known Activations