© 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. 26969
    Prev
    Next
    INDEX
    Explanations

    hexadecimal numbers and code prefixes

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

    The neuron fires on numeric constants—especially hexadecimal byte‐value literals (e.g. the “0x…” style 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
     old
    -1.05
    Through
    -0.99
    Since
    -0.98
    ensuremath
    -0.97
    Pří
    -0.96
     his
    -0.94
     konnte
    -0.94
     Ordens
    -0.92
    ꫀ
    -0.91
     same
    -0.91
    POSITIVE LOGITS
     nyingi
    1.11
    auml
    1.02
     baada
    1.01
    の人に
    1.01
    紧接着
    0.98
     agregó
    0.97
    是为了
    0.95
    cangan
    0.94
     canadian
    0.93
    跩
    0.93
    Activations Density 0.005%

    No Known Activations