© 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. 30492
    Prev
    Next
    INDEX
    Explanations

    retreating or moving back

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

    This neuron specializes in spotting LaTeX‐style mathematical notation (backslashes, braces, and math symbols) within the text.

    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
    躏
    -0.76
    (())
    -0.74
    的风
    -0.71
     ron
    -0.70
    🕎
    -0.69
    ährung
    -0.69
     بلو
    -0.69
    ulative
    -0.69
     Reduce
    -0.68
    celain
    -0.68
    POSITIVE LOGITS
     retreat
    4.72
     retreated
    4.00
     retreats
    3.98
     retreating
    3.86
     Retreat
    3.70
    退
    3.36
     backwards
    2.77
     backward
    2.73
     退
    2.58
    后退
    2.50
    Activations Density 0.051%

    No Known Activations