© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47387
    Prev
    Next
    INDEX
    Explanations

    every noun

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

    The neuron responds most strongly to rare or low‐frequency tokens (yielding high activations on uncommon words and lower activations on common ones).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     rokov
    -1.12
    色々
    -1.12
    整体
    -0.99
    様々
    -0.97
    ингл
    -0.96
    selves
    -0.94
     portions
    -0.92
    いろいろ
    -0.92
    rieren
    -0.92
    digt
    -0.91
    POSITIVE LOGITS
     imaginable
    1.87
     individually
    1.48
     وكل
    1.48
     except
    1.47
     separately
    1.44
     ทุก
    1.34
     every
    1.33
     that
    1.32
     einzeln
    1.30
    including
    1.17
    Activations Density 0.071%

    No Known Activations