© 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. 10-GEMMASCOPE-RES-131K
    7. 90394
    Prev
    Next
    INDEX
    Explanations

    network standards and addresses

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

    The neuron fires most strongly on the numeric segments and dots of dotted‐decimal IP addresses.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     tetrachloride
    -1.67
    </strong>
    -1.60
     qualquer
    -1.59
     juntamente
    -1.59
    rás
    -1.55
     quaisquer
    -1.53
     refirió
    -1.52
     limpiar
    -1.52
     crée
    -1.50
     serão
    -1.49
    POSITIVE LOGITS
    Rather
    1.95
    While
    1.80
    Since
    1.71
    Despite
    1.70
    Unlike
    1.68
    These
    1.66
    Although
    1.66
    Though
    1.62
    Those
    1.59
    Through
    1.53
    Activations Density 0.014%

    No Known Activations