© 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. 34-GEMMASCOPE-RES-131K
    7. 3217
    Prev
    Next
    INDEX
    Explanations

    natural, easy, second nature

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

    The neuron responds strongly to common, mid–to–high‐frequency English words (especially frequent verbs and function words), effectively acting as a word‐frequency detector.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ersen
    -0.89
    lery
    -0.80
    nos
    -0.75
    seits
    -0.74
    lean
    -0.74
    tison
    -0.73
     vigencia
    -0.72
    kapan
    -0.71
    eld
    -0.70
    rably
    -0.69
    POSITIVE LOGITS
     natural
    2.52
     second
    2.27
    natural
    2.06
    Natural
    1.88
     naturelle
    1.87
     easy
    1.85
     NATURAL
    1.77
     Natural
    1.77
     unnatural
    1.75
    second
    1.74
    Activations Density 0.025%

    No Known Activations