© 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. 27557
    Prev
    Next
    INDEX
    Explanations

    physical locations and objects

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

    The neuron detects words naming a domain or reference “surface” or “field” (e.g. field, ground, equator, ecliptic).

    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
     mellan
    -1.10
    ᧐
    -1.01
    ślę
    -0.99
     maravill
    -0.98
    ￦
    -0.93
     eenige
    -0.90
    )];
    -0.88
    WAUKEE
    -0.88
     между
    -0.84
    riel
    -0.84
    POSITIVE LOGITS
    譜
    1.00
    Other
    0.89
    deleted
    0.88
    Política
    0.85
    niss
    0.84
     lone
    0.84
    ۩
    0.84
     using
    0.83
    顛
    0.82
     Hanya
    0.82
    Activations Density 0.071%

    No Known Activations