© 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. 2572
    Prev
    Next
    INDEX
    Explanations

    equaled or tied records

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

    The neuron activates on numeric tokens and figures representing counts, scores, or statistics.

    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
     Bedien
    -0.85
     erwor
    -0.77
    zná
    -0.77
    ABULARY
    -0.75
     Huxley
    -0.75
    ᱐
    -0.75
    babkan
    -0.75
    体を
    -0.74
    riceps
    -0.74
     ense
    -0.73
    POSITIVE LOGITS
     matching
    2.80
     tying
    2.69
     equalled
    2.56
     equ
    2.48
     matched
    2.47
     tied
    2.45
    matching
    2.38
     Matching
    2.33
    Matching
    2.20
     equal
    2.19
    Activations Density 0.077%

    No Known Activations