© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 98464
    Prev
    Next
    INDEX
    Explanations

    why something happens

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

    The neuron fires on domain‐specific, content‐rich technical terms (e.g. “transparency,” “textual,” “mechanistic,” “understanding”) typically found in academic or scientific writing.

    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
    nių
    -0.84
     ángulo
    -0.81
    ösung
    -0.80
    कड
    -0.78
     patatas
    -0.77
    чів
    -0.77
    uun
    -0.76
     miroir
    -0.75
    áce
    -0.75
    jme
    -0.74
    POSITIVE LOGITS
     explanation
    1.95
     why
    1.70
     explain
    1.62
     explanations
    1.57
     explaining
    1.56
     reasons
    1.55
     explains
    1.52
    REASONS
    1.50
     WHY
    1.47
     Explanation
    1.39
    Activations Density 0.112%

    No Known Activations