© 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. 34-GEMMASCOPE-RES-131K
    7. 108936
    Prev
    Next
    INDEX
    Explanations

    spontaneous

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

    The neuron fires strongly on isolated single‐character tokens—especially non-Latin Unicode letters (e.g. Arabic or Greek letters) or hex digits in Unicode escape sequences.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    虽
    -0.90
     ہیں
    -0.88
    全程
    -0.85
    了两
    -0.84
    itosan
    -0.81
     keselamatan
    -0.81
    地に
    -0.81
    länge
    -0.79
     žaid
    -0.77
    .‏
    -0.77
    POSITIVE LOGITS
     XNUMX
    1.14
     ​​
    1.10
    never
    1.04
     incredibly
    1.02
    NEVER
    1.00
    ⓑ
    0.99
     dovr
    0.96
    implies
    0.96
    かっこ
    0.95
    typically
    0.94
    Activations Density 0.000%

    No Known Activations