© 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. 15418
    Prev
    Next
    INDEX
    Explanations

    everything or all

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

    The neuron fires on the document’s key, content‐bearing words—especially longer, topic-specific nouns or verbs—i.e. the main technical or subject terms of each passage.

    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
    userRepository
    -0.91
    čeno
    -0.86
    /'.$
    -0.85
     because
    -0.85
     on
    -0.83
    рения
    -0.83
     algún
    -0.81
    apatos
    -0.81
     in
    -0.79
     by
    -0.79
    POSITIVE LOGITS
     everything
    2.30
     EVERYTHING
    1.89
    everything
    1.73
    Everything
    1.39
     både
    1.34
     تمامی
    1.32
    ridescent
    1.32
     segala
    1.27
    あらゆる
    1.27
     everywhere
    1.26
    Activations Density 0.228%

    No Known Activations