© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101319
    Prev
    Next
    INDEX
    Explanations

    emphasizes importance

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

    The neuron fires strongly on words and subword pieces that mean “to emphasize” or “to stress” (e.g. stressed, stressing, emphasise, emphasising, pabrėžti subparts).

    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
     Such
    -1.82
    <
    -1.78
    ...
    -1.64
     imaginar
    -1.62
    (
    -1.58
     Other
    -1.56
     �
    -1.55
    ",
    -1.55
     damn
    -1.53
    —
    -1.52
    POSITIVE LOGITS
     importance
    2.13
     będziesz
    2.00
    𖥦
    1.90
     Konkur
    1.72
     musisz
    1.70
    涠
    1.66
     będ
    1.65
    喼
    1.65
    which
    1.63
     chcesz
    1.61
    Activations Density 0.030%

    No Known Activations