© 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. 66071
    Prev
    Next
    INDEX
    Explanations

    what follows it

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

    This neuron fires on common English “function” words—short closed-class items like conjunctions (and, though, which, so), pronouns (it, I, they), and articles/determiners (a).

    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
    為に
    -2.48
    ほうがいい
    -2.25
    ところに
    -2.23
    ほうが
    -2.11
    を探す
    -2.06
    感じで
    -2.05
    ようになりました
    -1.92
    様な
    -1.88
     hés
    -1.88
    方がいい
    -1.86
    POSITIVE LOGITS
    を
    2.55
     also
    2.38
    が
    2.30
    や
    2.16
    は
    2.08
     in
    2.03
    も
    2.02
    と
    1.98
    がある
    1.84
     even
    1.77
    Activations Density 0.100%

    No Known Activations