© 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. 22-GEMMASCOPE-RES-131K
    7. 65222
    Prev
    Next
    INDEX
    Explanations

    questions about why

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

    The neuron fires on discourse-level cue phrases and question words that signal sentence introductions or transitions.

    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
    ::
    -0.94
     плит
    -0.86
     ”
    -0.85
     Genuine
    -0.85
     j
    -0.84
    ტი
    -0.84
    لسط
    -0.84
    stdc
    -0.83
     sehen
    -0.83
    zesi
    -0.82
    POSITIVE LOGITS
    elegante
    1.15
    っていますが
    1.13
     lomba
    1.13
     подтвер
    1.06
     magasiner
    1.04
    emporio
    1.02
    ressa
    1.00
    grees
    1.00
    produksi
    0.98
     vider
    0.98
    Activations Density 0.077%

    No Known Activations