© 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. 99936
    Prev
    Next
    INDEX
    Explanations

    FITNESS FOR A PARTICULAR PURPOSE

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

    The neuron fires on numeric literals, especially floating‐point numbers (decimal 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_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
    Because
    -2.97
    There
    -2.97
    One
    -2.72
    In
    -2.69
    Almost
    -2.66
    Not
    -2.61
    Many
    -2.61
    If
    -2.59
    All
    -2.59
    𖥦
    -2.58
    POSITIVE LOGITS
    enablog
    3.06
    閙
    3.05
    轳
    3.02
    憷
    3.00
    𖤍
    2.98
     hefty
    2.88
    粜
    2.84
    
    2.81
     rivales
    2.80
    擡
    2.80
    Activations Density 0.002%

    No Known Activations