© 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. 22-GEMMASCOPE-RES-131K
    7. 27801
    Prev
    Next
    INDEX
    Explanations

    users and roles

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

    This neuron responds strongly to mentions of user or user-type terms, i.e. references to different user roles or audiences (e.g. “users,” “end-users,” “power-user,” “developer,” “biologist,” 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
     antiguas
    -1.23
     this
    -1.22
    简约
    -1.21
     zahlreichen
    -1.09
     animals
    -1.08
     ingredienti
    -1.05
    حن
    -1.05
     Grunds
    -1.04
     vecchia
    -1.03
     corrientes
    -1.02
    POSITIVE LOGITS
     users
    1.70
     who
    1.61
     bedien
    1.47
     besoins
    1.33
     rédu
    1.24
     user
    1.24
     பய
    1.23
    Use
    1.23
    Uses
    1.21
     infeksi
    1.21
    Activations Density 0.089%

    No Known Activations