© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 21595
    Prev
    Next
    INDEX
    Explanations

    completion or end

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

    The neuron fires on words (and substrings) that signify “end” or “completion” (e.g. finished/finished, end, completed, 終了, the Russian root кон).

    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
    geſ
    -2.75
    during
    -2.34
    bbero
    -2.23
    enthusi
    -2.23
     THROUGH
    -2.20
    through
    -2.20
    ⧠
    -2.19
    秶
    -2.17
     através
    -2.14
    锇
    -2.09
    POSITIVE LOGITS
     With
    3.02
    S
    2.75
    end
    2.30
    T
    2.30
     =
    2.22
     Currently
    2.20
    ～～
    2.14
    O
    2.11
    ma
    2.11
     These
    2.08
    Activations Density 0.006%

    No Known Activations