© 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. 22-GEMMASCOPE-RES-131K
    7. 67059
    Prev
    Next
    INDEX
    Explanations

    single letters followed by acronyms

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

    This neuron strongly activates on isolated uppercase letter tokens (e.g. single-letter headings or acronyms).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     at
    -2.75
     epitom
    -2.67
     ostensibly
    -2.56
     with
    -2.52
     in
    -2.38
     blatantly
    -2.34
    ！！！
    -2.31
     all
    -2.31
     has
    -2.30
     first
    -2.28
    POSITIVE LOGITS
    蓐
    2.88
    ⁛
    2.69
    ܆
    2.64
    艄
    2.61
    の使用
    2.61
    
    2.56
    に入れる
    2.53
    dersfield
    2.50
    阏
    2.48
    厹
    2.48
    Activations Density 0.613%

    No Known Activations