© 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. 67253
    Prev
    Next
    INDEX
    Explanations

    last followed by a sequence

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

    The neuron is highly responsive to occurrences of the word “last” (in any case).

    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
    綃
    -1.82
    了自己的
    -1.78
     COMES
    -1.76
     abbiamo
    -1.69
     لعبة
    -1.68
    appeared
    -1.68
    weren
    -1.68
    峁
    -1.67
    came
    -1.66
    allows
    -1.65
    POSITIVE LOGITS
     всех
    1.84
     could
    1.84
    尤其
    1.81
     possibly
    1.68
     would
    1.65
     also
    1.64
     continuously
    1.63
     fortæ
    1.59
     after
    1.55
    gelijk
    1.54
    Activations Density 0.016%

    No Known Activations