© 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. 59535
    Prev
    Next
    INDEX
    Explanations

    company names and code identifiers

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

    The neuron responds to tokens written in uppercase (or largely uppercase) form, such as acronyms, code labels, or proper-name fragments in all caps.

    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
    Ꮉ
    -2.92
    沒想到
    -2.34
    不
    -2.31
    
    -2.30
    鹪
    -2.30
    亖
    -2.27
    猡
    -2.25
    嘖
    -2.25
    鱘
    -2.23
    
    -2.22
    POSITIVE LOGITS
    ’
    3.38
    甚至
    3.22
    as
    3.11
    re
    2.88
    er
    2.80
    或
    2.80
    可以
    2.70
    ar
    2.66
    和
    2.47
    one
    2.44
    Activations Density 0.008%

    No Known Activations