© 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. 72739
    Prev
    Next
    INDEX
    Explanations

    associated with

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

    This neuron fires on content‐bearing words (nouns, adjectives, verbs) rather than on common function words.

    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
    umque
    -0.91
     erreurs
    -0.88
     ഉ
    -0.86
    とも言
    -0.84
     качества
    -0.84
     dahil
    -0.82
     علا
    -0.82
    すっ
    -0.82
     aussieht
    -0.81
    IFORNIA
    -0.80
    POSITIVE LOGITS
     these
    1.54
     this
    1.46
     sometimes
    1.33
     ilyen
    1.14
     now
    1.13
     evidently
    1.13
     décid
    1.12
    就是这样
    1.11
     truly
    1.09
     verkligen
    1.09
    Activations Density 0.047%

    No Known Activations