© 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. 10-GEMMASCOPE-RES-131K
    4. 99666
    Prev
    Next
    INDEX
    Explanations

    code loops

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

    This neuron never fires—it's effectively a dead neuron that doesn’t detect any pattern.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     and
    -3.47
     as
    -3.14
     There
    -2.86
     to
    -2.86
     so
    -2.63
     or
    -2.44
     What
    -2.42
     on
    -2.41
     in
    -2.38
     In
    -2.38
    POSITIVE LOGITS
     diki
    3.22
     cinto
    3.08
     elektrik
    2.95
     hunde
    2.88
     entusias
    2.84
    2
    2.83
     keramik
    2.73
     seksi
    2.72
    0
    2.72
     komik
    2.70
    Activations Density 0.003%

    No Known Activations