© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 71224
    Prev
    Next
    INDEX
    Explanations

    negative numbers

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

    The neuron activates on numeric literals (especially signed or multi‐digit numbers) in code and math expressions.

    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
    -1.84
     Its
    -1.83
     then
    -1.73
     You
    -1.67
     县
    -1.53
    等は
    -1.52
     现场
    -1.51
    幃
    -1.49
    We
    -1.48
     '
    -1.48
    POSITIVE LOGITS
    
    1.94
     Bacteriol
    1.94
    蔟
    1.91
     fossero
    1.84
     gustaba
    1.81
    繧
    1.77
     avantaj
    1.71
    鷽
    1.70
     kolye
    1.65
     avevano
    1.65
    Activations Density 0.014%

    No Known Activations