© 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. 38271
    Prev
    Next
    INDEX
    Explanations

    html tag attributes like class

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

    The neuron is triggered by tokens containing numeric characters (e.g. digits in IDs, CSS sizes, or other numeric values).

    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
     herself
    -2.05
     {
    -1.15
     город
    -1.09
     która
    -1.09
    esh
    -1.06
    dding
    -1.05
     the
    -1.02
     которая
    -1.02
     away
    -1.01
     Unterricht
    -1.01
    POSITIVE LOGITS
     himself
    1.66
     نفسه
    1.37
     necessari
    1.34
    duje
    1.17
     ?',
    1.16
    ācijas
    1.16
     енер
    1.14
     necesar
    1.14
    inę
    1.14
     الداخلية
    1.13
    Activations Density 0.021%

    No Known Activations