© 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. 34-GEMMASCOPE-RES-131K
    7. 107884
    Prev
    Next
    INDEX
    Explanations

    to proceed

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

    This neuron primarily fires on the token “to,” especially when it marks the start of an infinitive or an instruction (e.g., “to remove,” “to cancel,” “to open”).

    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
    Provide
    -1.09
    provide
    -1.06
     providing
    -0.92
     Provide
    -0.90
    provides
    -0.90
     fournir
    -0.90
    了啊
    -0.88
     gewisse
    -0.88
    <bos>
    -0.87
     those
    -0.87
    POSITIVE LOGITS
     see
    1.63
     begin
    1.34
    begin
    1.24
     your
    1.23
    reiches
    1.23
    yourself
    1.18
     mulai
    1.18
    เริ่ม
    1.16
    开始
    1.15
     yourself
    1.11
    Activations Density 0.049%

    No Known Activations