© 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. 112708
    Prev
    Next
    INDEX
    Explanations

    contractions and possessives

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

    The neuron is principally triggered by tokens containing an apostrophe (e.g. in contractions or possessives).

    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
    Beskrivning
    -1.60
    你不要
    -1.39
    取締
    -1.38
     Some
    -1.35
    Цита
    -1.34
    ઃ
    -1.34
    に行ってきました
    -1.30
    柠檬
    -1.29
     have
    -1.27
    Takk
    -1.26
    POSITIVE LOGITS
    <bos>
    2.17
    </i>
    1.75
    </em>
    1.67
    ين
    1.66
     diminuer
    1.52
     vacunas
    1.51
     gefahren
    1.48
    ljeno
    1.47
    undidad
    1.46
    EMBLY
    1.45
    Activations Density 0.043%

    No Known Activations