© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 54020
    Prev
    Next
    INDEX
    Explanations

    codes and symbols

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

    This neuron activates on isolated mathematical or technical tokens—e.g. single‐letter variables, LaTeX commands (nabla, mathbf), underscores, or other formula parts.

    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
     Matériau
    -1.16
    culously
    -1.14
    ensively
    -1.12
     Mejora
    -1.03
    に来
    -1.02
     botellas
    -1.01
    %",
    -1.00
    苻
    -1.00
    を探す
    -1.00
    にしました
    -1.00
    POSITIVE LOGITS
     of
    1.88
    gangenheit
    0.85
    et
    0.84
    спорта
    0.82
    である
    0.80
    to
    0.79
    Ỡ
    0.78
    level
    0.77
     hulle
    0.77
     atve
    0.76
    Activations Density 0.049%

    No Known Activations