© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 39217
    Prev
    Next
    INDEX
    Explanations

    code block delimiters

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

    The neuron fires strongly on numeric literals (numbers) 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
     baie
    -1.16
     ADMIN
    -1.00
    doors
    -1.00
     Secondo
    -0.93
     gut
    -0.93
    kjø
    -0.92
    BarButtonItem
    -0.91
    ❣️
    -0.90
     bani
    -0.90
    加热
    -0.89
    POSITIVE LOGITS
    这种情况
    1.14
    While
    1.09
    Though
    1.05
     notwendi
    1.01
    After
    0.89
     praktischen
    0.88
    感じです
    0.86
    ينما
    0.86
    の様子
    0.86
    Despite
    0.86
    Activations Density 0.068%

    No Known Activations