© 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. 22-GEMMASCOPE-RES-131K
    7. 128184
    Prev
    Next
    INDEX
    Explanations

    The neuron broadly responds to very common high-frequency tokens (mostly function words, punctuation, and other frequent vocabulary).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    only one

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     fabrik
    -0.89
    帘
    -0.87
     pregi
    -0.85
     集合
    -0.84
    veal
    -0.84
     sintomi
    -0.84
    िका
    -0.81
     particolarmente
    -0.81
     risol
    -0.81
     rilev
    -0.80
    POSITIVE LOGITS
    except
    1.09
     only
    1.05
    variables
    1.04
    唯一
    1.03
     one
    1.02
    倒是
    1.01
    变化
    1.00
     constants
    0.98
     excepto
    0.96
     zůst
    0.96
    Activations Density 0.054%

    No Known Activations