© 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. 10-GEMMASCOPE-RES-131K
    7. 70864
    Prev
    Next
    INDEX
    Explanations

    Geoff, conform, demo, Ω

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

    The neuron responds to uncommon or specialized tokens—like rare proper names, Greek symbols, or unique class/package identifiers.

    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
    AutoresizingMask
    -2.55
    蹋
    -2.42
    ......
    -2.34
     emphatic
    -2.17
    </h3>
    -2.16
    蹰
    -2.16
     organize
    -2.14
    .....
    -2.14
    大陸
    -2.14
     شوند
    -2.13
    POSITIVE LOGITS
    ers
    4.38
    鯪
    3.78
    n
    3.36
    {
    3.33
    𓐍
    2.61
    羮
    2.56
     berdua
    2.53
    gete
    2.53
    arians
    2.52
    𓆏
    2.52
    Activations Density 0.009%

    No Known Activations