© 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. 22-GEMMASCOPE-RES-131K
    7. 38631
    Prev
    Next
    INDEX
    Explanations

    determiner followed by content type

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

    The neuron activates on words that signal relevance or point to content—e.g. demonstratives (“this,” “that”) and relevance markers (“informative,” “relevant”).

    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
    hauled
    -1.05
     τῶν
    -0.99
    embl
    -0.99
    ङ
    -0.97
    batsman
    -0.97
    emate
    -0.96
     guapos
    -0.96
    forhold
    -0.95
     who
    -0.94
     tüm
    -0.93
    POSITIVE LOGITS
     article
    1.29
     lunghezza
    1.11
     that
    1.07
     fornito
    1.07
     uska
    1.05
    providedIn
    1.05
     капю
    1.04
    borderSide
    1.02
    sigkeit
    1.01
    ďaka
    1.00
    Activations Density 0.048%

    No Known Activations