© 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. 59770
    Prev
    Next
    INDEX
    Explanations

    All [X] [Y]

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

    The neuron is specialized to spot mentions of formal organization names or their acronyms (e.g. “All India Football Federation,” “Panhellenic,” “AIR,” etc.).

    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
    \
    -1.88
     has
    -1.77
    P
    -1.76
    O
    -1.66
    หรือ
    -1.65
     then
    -1.65
    T
    -1.65
    F
    -1.65
    In
    -1.59
    S
    -1.57
    POSITIVE LOGITS
    vimento
    1.80
    resten
    1.67
     minute
    1.62
    ,
    1.59
    
    1.57
    ﻂ
    1.57
    狁
    1.55
    ."'
    1.55
    瘰
    1.55
    itsubishi
    1.54
    Activations Density 0.012%

    No Known Activations