© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72623
    Prev
    Next
    INDEX
    Explanations

    desire for things and actions

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

    This neuron picks out the specific thing a user is asking for—that is, the noun or phrase naming the object of an “I want…”/request sentence.

    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
     using
    -1.09
     répondu
    -1.08
    lepší
    -1.05
     felé
    -1.05
    還有
    -1.04
    blusa
    -1.02
     allowing
    -1.02
     needing
    -1.02
     garantizar
    -1.01
    Выбор
    -1.00
    POSITIVE LOGITS
     to
    2.78
    erving
    1.07
     unbedingt
    1.04
    vonne
    1.02
    тата
    1.02
    setColumns
    0.98
     להיות
    0.98
    ONY
    0.98
    vano
    0.97
    rowa
    0.97
    Activations Density 0.059%

    No Known Activations