© 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. 64948
    Prev
    Next
    INDEX
    Explanations

    what clauses

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

    The neuron detects occurrences of the phrase “know what … are/you are doing.”

    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
     have
    -1.49
     colgar
    -1.22
     publicação
    -1.21
     also
    -1.19
    Jadi
    -1.13
    了解更多
    -1.12
     ability
    -1.11
     that
    -1.08
     kunnen
    -1.06
     cristianismo
    -1.06
    POSITIVE LOGITS
     to
    1.68
     sınır
    1.02
     where
    1.01
     wohin
    1.00
    рованию
    1.00
     whofe
    0.96
    </td>
    0.96
    </h5>
    0.96
     preserv
    0.93
     worauf
    0.92
    Activations Density 0.017%

    No Known Activations