© 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. 34-GEMMASCOPE-RES-131K
    7. 73610
    Prev
    Next
    INDEX
    Explanations

    sharp

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

    The neuron strongly activates on the character sequence “sharp,” i.e. it detects occurrences of the word (or name) “sharp” in any context.

    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
    💼
    -0.88
     absoluta
    -0.79
     tamanho
    -0.79
    affected
    -0.79
     esforço
    -0.78
    台中
    -0.77
    轧
    -0.77
    筋肉
    -0.77
    asist
    -0.77
    aan
    -0.76
    POSITIVE LOGITS
     sharp
    1.73
    sharp
    1.68
    Sharp
    1.61
     Sharp
    1.59
     SHARP
    1.33
     edges
    1.32
     sharpness
    1.32
     sharper
    1.30
     knives
    1.12
     razor
    1.11
    Activations Density 0.019%

    No Known Activations