© 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. 22-GEMMASCOPE-RES-131K
    7. 14974
    Prev
    Next
    INDEX
    Explanations

    design studio ; 0 px FIFA /* D Q

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

    The neuron fires whenever it sees numeric literal tokens (e.g. CSS position or dimension values) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    therners
    -1.21
     gwaran
    -1.18
     habang
    -1.15
     kapag
    -1.13
     wspania
    -1.08
    えっ
    -1.07
     princesas
    -1.06
    ofing
    -1.05
     ако
    -1.02
    clientWidth
    -1.02
    POSITIVE LOGITS
    面積
    0.95
    結成
    0.94
     other
    0.89
     chert
    0.89
    などの
    0.88
     gaunt
    0.84
     также
    0.84
    вах
    0.84
    というのも
    0.84
     showcasing
    0.83
    Activations Density 0.005%

    No Known Activations