© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 70145
    Prev
    Next
    INDEX
    Explanations

    places and technical terms

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

    The neuron activates on capitalized proper‐name tokens (e.g. place names like “Essex” or acronyms like “SPDY”).

    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
    ↵
    -3.67
    淠
    -2.89
    鳢
    -2.81
    鲮
    -2.73
    囖
    -2.69
    媗
    -2.69
    毉
    -2.66
    蟜
    -2.64
    嫱
    -2.64
     短发
    -2.64
    POSITIVE LOGITS
    t
    3.59
    en
    3.08
    s
    2.94
    d
    2.89
    к
    2.67
    om
    2.63
    𝑚
    2.59
     пев
    2.58
    2.53
    ?"
    2.53
    Activations Density 0.016%

    No Known Activations