© 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. 42678
    Prev
    Next
    INDEX
    Explanations

    glistening sweetie FLD Khu

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

    The neuron responds to uncommon or out-of-context tokens—e.g. proper names, acronyms, loanwords, or striking descriptive words.

    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
    翃
    -3.03
    e
    -2.98
    鲀
    -2.86
     самая
    -2.81
     unsere
    -2.67
    ⤥
    -2.64
     глав
    -2.58
    o
    -2.48
    𖤍
    -2.45
     This
    -2.42
    POSITIVE LOGITS
    寕
    2.83
    鲎
    2.70
    蹋
    2.64
    を行い
    2.61
    厖
    2.53
    ’
    2.50
    筇
    2.45
    gucig
    2.44
    ዡ
    2.42
    瘩
    2.31
    Activations Density 0.001%

    No Known Activations