© 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. 34-GEMMASCOPE-RES-131K
    7. 32936
    Prev
    Next
    INDEX
    Explanations

    graph and logic components

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

    The neuron fires on highly technical, domain-specific academic jargon—especially specialized mathematical and theoretical-computer-science terms (e.g. “complexity,” “equivalence,” “automata,” “polynomial identity testing,” “similarity-based objective functions,” etc.).

    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
     présentent
    -0.93
    摟
    -0.93
    лении
    -0.90
     alterar
    -0.86
     送
    -0.85
     mustache
    -0.82
     باشی
    -0.81
     frete
    -0.81
     tampaknya
    -0.81
    Epidemiology
    -0.81
    POSITIVE LOGITS
     semantics
    0.93
     aware
    0.92
     in
    0.89
     graphs
    0.88
    protected
    0.85
    dba
    0.83
    lect
    0.79
    Vec
    0.79
    based
    0.79
    znym
    0.79
    Activations Density 0.105%

    No Known Activations