© 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. 87349
    Prev
    Next
    INDEX
    Explanations

    a better adjective

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

    This neuron fires on high–information, content‐bearing tokens (i.e. words with high tf–idf/informativeness).

    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
     of
    -1.74
    Palabras
    -1.40
     prachtige
    -1.38
    却又
    -1.34
     belangrijke
    -1.30
     önem
    -1.30
     keduanya
    -1.27
     Garantía
    -1.26
     Sekarang
    -1.26
     bijzondere
    -1.25
    POSITIVE LOGITS
     which
    1.84
     first
    1.71
     Why
    1.27
    You
    1.27
     är
    1.27
     What
    1.24
    They
    1.21
     what
    1.16
     Those
    1.13
     голове
    1.13
    Activations Density 0.096%

    No Known Activations