© 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. 10-GEMMASCOPE-RES-131K
    7. 12799
    Prev
    Next
    INDEX
    Explanations

    generate instructions

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

    The neuron fires most strongly on longer, content-heavy (i.e. non–stopword) tokens—particularly rare or multi-syllabic terms—regardless of domain.

    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
    -2.02
     has
    -1.84
     first
    -1.80
     may
    -1.78
     will
    -1.76
     gdyż
    -1.73
     always
    -1.62
     In
    -1.61
    ?
    -1.59
     five
    -1.58
    POSITIVE LOGITS
    こない
    1.99
     Scritto
    1.98
     baumwolle
    1.97
     ſont
    1.85
    diese
    1.83
     pittores
    1.82
     zoude
    1.81
     camiset
    1.77
     jezt
    1.76
    ㍉
    1.70
    Activations Density 22.891%

    No Known Activations