© 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. 73761
    Prev
    Next
    INDEX
    Explanations

    Downing Street, Eskom, extrinsic, moons, downs

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

    The neuron fires on occurrences of the substring “downs” (e.g. the word “downs”).

    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
    What
    -3.31
    Since
    -3.28
    Also
    -3.11
    Initially
    -2.97
    Like
    -2.94
     This
    -2.84
    Not
    -2.83
     Actually
    -2.81
    Perhaps
    -2.81
    During
    -2.78
    POSITIVE LOGITS
                    
    2.69
    蠖
    2.69
    ){
    2.67
    祧
    2.58
    m
    2.50
    采用了
    2.48
    澌
    2.44
    髫
    2.41
    ↵↵
    2.38
    ſſen
    2.36
    Activations Density 0.006%

    No Known Activations