© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 70475
    Prev
    Next
    INDEX
    Explanations

    reclusive, mysterious, seldom seen

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

    The neuron flags words that convey secrecy, exclusivity, or hiddenness (e.g. selective, reclusive, cryptic, mysterious, shadowy).

    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
     already
    -0.99
    าก็
    -0.94
     уже
    -0.94
     sentimientos
    -0.93
     then
    -0.87
     astronomer
    -0.84
     bereits
    -0.83
    でもある
    -0.82
    uyler
    -0.82
     ストレッチ
    -0.82
    POSITIVE LOGITS
     rarely
    1.52
     seldom
    1.35
    quiet
    1.34
     quietly
    1.34
     infrequently
    1.30
     quiet
    1.25
    Quiet
    1.23
    isHidden
    1.17
     hardly
    1.15
    Unlike
    1.14
    Activations Density 0.055%

    No Known Activations