© 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. 15371
    Prev
    Next
    INDEX
    Explanations

    when something happens

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

    The neuron detects temporal‐trigger phrases—especially the word “when” introducing time or conditional clauses.

    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
     it
    -1.48
     same
    -1.32
     if
    -1.23
    s
    -1.23
     {
    -1.19
     into
    -1.17
     metales
    -1.17
    te
    -1.16
     considerations
    -1.15
    但她
    -1.15
    POSITIVE LOGITS
    исправ
    1.61
    ︓
    1.53
     a
    1.47
    тата
    1.45
    heureux
    1.41
    libatkan
    1.38
    batas
    1.36
    стандарт
    1.32
    álise
    1.30
     ペット
    1.29
    Activations Density 0.027%

    No Known Activations