© 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. 39322
    Prev
    Next
    INDEX
    Explanations

    asking why or for what reason

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

    The neuron detects occurrences of causal question words—especially “why” (and the accompanying question contexts).

    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
    hawatir
    -0.94
     romantique
    -0.92
     uji
    -0.90
    itaan
    -0.88
    carian
    -0.88
     heartbreak
    -0.86
     diinginkan
    -0.86
     Überras
    -0.86
    robat
    -0.86
    candidate
    -0.86
    POSITIVE LOGITS
     why
    1.48
     Why
    1.36
    Why
    1.18
     ऐसा
    1.16
    why
    1.16
    这种
    1.15
     waarom
    1.13
     Such
    1.09
    Such
    1.05
    之所以
    1.05
    Activations Density 0.077%

    No Known Activations