© 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. 130046
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numerical automotive specification details (e.g. seat counts, MPG figures, wheel/drive descriptors) in car ads.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    "rebounds" or "designed" or "world record"

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     only
    -1.46
     on
    -1.33
     because
    -1.26
     one
    -1.18
     several
    -1.17
     five
    -1.16
     despite
    -1.15
     epitom
    -1.13
     an
    -1.13
     at
    -1.10
    POSITIVE LOGITS
    Leírás
    1.32
     nonchal
    1.24
    っており
    1.19
    Παρα
    1.17
     jednod
    1.17
     Dinas
    1.16
     También
    1.16
     çeşit
    1.16
    somewhat
    1.16
     begrij
    1.15
    Activations Density 0.087%

    No Known Activations