© 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. 270
    Prev
    Next
    INDEX
    Explanations

    numbers and units like .com or .ft

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

    numerical data in structured table and list formats.

    oai_token-act-pair · claude-4-5-haikuTriggered by @jyhe0408

    The neuron fires on numeric tokens in structured tables or lists—especially draw or standings numbers in sports‐results pages.

    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
     概要
    -1.10
     effectu
    -1.05
     Atau
    -1.00
     Ahora
    -0.98
    沵
    -0.96
     manufac
    -0.95
     augusti
    -0.93
     BEAUT
    -0.93
    rethe
    -0.93
    ⅿ
    -0.92
    POSITIVE LOGITS
    しましたが
    1.09
     Additional
    1.06
     additional
    0.98
     uses
    0.98
     their
    0.97
     Gemein
    0.96
    opheles
    0.96
     what
    0.94
    erobic
    0.94
    frances
    0.94
    Activations Density 0.013%

    No Known Activations