© 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. 10-GEMMASCOPE-RES-131K
    7. 72671
    Prev
    Next
    INDEX
    Explanations

    According to

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

    The neuron strongly activates on the very first word of a new paragraph or standalone heading (i.e. tokens that begin a text block).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     which
    -1.45
     how
    -1.45
     that
    -1.43
    早晨
    -1.41
     those
    -1.30
     ktor
    -1.30
    iously
    -1.29
    onActivityResult
    -1.29
    glPopMatrix
    -1.29
    ERÍA
    -1.27
    POSITIVE LOGITS
    ➴
    1.38
     Geografia
    1.34
     även
    1.30
     on
    1.29
     perpé
    1.29
     Kühlschrank
    1.25
    と感じ
    1.22
    也好
    1.20
     diriger
    1.19
    <td>
    1.17
    Activations Density 0.109%

    No Known Activations