© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65728
    Prev
    Next
    INDEX
    Explanations

    abilities and qualities

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

    The neuron fires on words denoting positive personal attributes or abilities (e.g. talents, spirit, voice, songwriting, versatility).

    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
     themselves
    -1.03
     sienten
    -1.02
    themselves
    -1.02
    彼らは
    -0.99
     verfügen
    -0.98
    нє
    -0.98
     it
    -0.98
    доступ
    -0.98
    volve
    -0.97
    volves
    -0.97
    POSITIVE LOGITS
     and
    1.22
     help
    1.20
     continue
    1.05
     เห
    1.03
     njego
    1.02
    euvre
    1.00
     survive
    0.95
     durante
    0.94
    と思われる
    0.94
    cinct
    0.93
    Activations Density 0.028%

    No Known Activations