© 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. 127654
    Prev
    Next
    INDEX
    Explanations

    I just

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

    The neuron is detecting and responding to numeric tokens (i.e. numbers) in 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_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
     to
    -2.02
     also
    -1.93
     will
    -1.83
     its
    -1.70
    
    -1.53
     juos
    -1.52
     last
    -1.51
    一首
    -1.45
     "@
    -1.44
    :][
    -1.42
    POSITIVE LOGITS
    larda
    1.70
     klasik
    1.66
     relâche
    1.62
     katk
    1.61
     parfü
    1.55
     kapak
    1.53
    Hacer
    1.52
    mientras
    1.52
     rajut
    1.48
     senin
    1.46
    Activations Density 0.039%

    No Known Activations