© 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
    2. Gemma-3-27B-IT
    3. 3-GEMMASCOPE-2-TRANSCODER-262K
    4. 46479
    Prev
    Next
    INDEX
    Explanations

    Based on the provided data, the neuron seems to be primarily associated with concepts of rounding numerical values, often to a specific number of decimal places. The presence of "round off," "round it off," "round it up," and "two places" in the activating texts strongly suggests this.Therefore, the most fitting explanation is:**round off numeric values**round off numeric values

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_3_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    es
    0.80
    c
    0.78
    س
    0.75
    Ring
    0.71
    स
    0.69
    cA
    0.68
    ه
    0.68
    ات
    0.66
    octagon
    0.65
    cenie
    0.65
    POSITIVE LOGITS
     cataly
    0.67
     reorgan
    0.63
    ని
    0.57
     legitim
    0.57
    اد
    0.55
     synchron
    0.55
     galvan
    0.54
     sensory
    0.54
     appease
    0.54
    ного
    0.54
    Activations Density 0.003%

    No Known Activations