© 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. 15372
    Prev
    Next
    INDEX
    Explanations

    SVG path data and coordinates

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

    The neuron primarily activates on numeric tokens (digits or numbers) in technical/xml/code contexts.

    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
    Exemple
    -1.06
    begins
    -1.05
     who
    -1.02
    vriende
    -1.02
    happened
    -1.00
    weren
    -0.99
     σω
    -0.99
     bluse
    -0.96
     barbati
    -0.94
    却没有
    -0.94
    POSITIVE LOGITS
     and
    1.25
     by
    1.03
     while
    0.96
     other
    0.94
     we
    0.92
    miel
    0.91
     five
    0.91
    同志
    0.89
     rebellion
    0.88
    misión
    0.85
    Activations Density 0.005%

    No Known Activations