© 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. 34-GEMMASCOPE-RES-131K
    7. 90690
    Prev
    Next
    INDEX
    Explanations

    want to verb

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

    The neuron detects occurrences of the phrase “want to,” highlighting where the text expresses a user’s desire or intent.

    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
     możliwość
    -1.09
     vestito
    -1.09
    資料來源
    -1.03
    ⧠
    -1.02
     merkwür
    -1.02
     blaue
    -1.02
    miesz
    -1.01
    gantung
    -1.01
    ‴
    -1.01
    ᐠ
    -1.00
    POSITIVE LOGITS
     to
    2.20
    的是
    1.12
     know
    1.09
     receive
    1.06
     from
    1.02
     a
    1.02
     an
    0.98
     видеть
    0.96
     really
    0.96
     get
    0.90
    Activations Density 0.089%

    No Known Activations