© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 130439
    Prev
    Next
    INDEX
    Explanations

    The neuron reliably picks out numeric values (including integers, decimals, percentages) and other measurement‐style or parameter tokens (e.g. “function,” “mode,” “axes,” “LIKE”) that typically appear as data entries or key technical terms in tables and code snippets.

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

    structure or definition

    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
     plafon
    -1.08
     rapides
    -1.06
    LOCK
    -1.04
    ครื่อง
    -1.02
     conçus
    -1.01
    wachen
    -0.99
     konfer
    -0.99
    Keyboard
    -0.97
    csak
    -0.97
    шите
    -0.96
    POSITIVE LOGITS
     from
    1.17
     at
    1.14
     three
    1.02
     first
    1.02
     by
    1.00
     not
    0.96
     unbelievable
    0.94
     there
    0.92
     two
    0.92
     beginning
    0.90
    Activations Density 0.046%

    No Known Activations