© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28462
    Prev
    Next
    INDEX
    Explanations

    pleaded guilty to accident

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

    The neuron responds to non-word tokens—especially punctuation marks like apostrophes and periods and numeric digits.

    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
     innebär
    -0.92
    我可以
    -0.91
     Bue
    -0.83
     eignen
    -0.79
    зит
    -0.78
     phrasing
    -0.75
    داية
    -0.75
    hasOwnProperty
    -0.75
    Са
    -0.74
     compliqué
    -0.74
    POSITIVE LOGITS
    inazione
    0.92
    hian
    0.85
    ramine
    0.84
     ناز
    0.84
    ɠ
    0.82
    segmentation
    0.81
    essence
    0.80
    ٖ
    0.80
    鲱
    0.79
    cession
    0.79
    Activations Density 0.002%

    No Known Activations