© 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. 28599
    Prev
    Next
    INDEX
    Explanations

    state of function or application

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

    The neuron fires on language describing a technical assumption or method “failing” or “breaking down,” i.e. phrases that signal an approximation losing validity.

    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
     but
    -0.93
     rather
    -0.92
     medically
    -0.87
     députés
    -0.87
     crowd
    -0.86
    老板
    -0.85
     crowds
    -0.85
    格式
    -0.85
    ковь
    -0.85
     nowadays
    -0.83
    POSITIVE LOGITS
    eloma
    1.10
    performs
    1.02
    </>
    1.02
    itisation
    0.97
    warranty
    0.96
     its
    0.96
     εφαρ
    0.96
     dégustation
    0.95
     applied
    0.95
    ibile
    0.94
    Activations Density 0.231%

    No Known Activations