© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 36031
    Prev
    Next
    INDEX
    Explanations

    When followed by risk, Q followed by :, the followed by goal

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

    This neuron detects URL‐style percent‐encoded character sequences (e.g. “%D8%B5”).

    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
    L
    -1.35
    A
    -1.29
    P
    -1.28
    R
    -1.21
    t
    -1.08
    ,
    -1.05
    C
    -1.05
    D
    -1.04
     Don
    -1.03
    様な
    -1.03
    POSITIVE LOGITS
    ܧ
    1.35
     botines
    1.25
     ちゃ
    1.22
     criatura
    1.19
     escultura
    1.18
     atencion
    1.17
    Initializing
    1.17
     manzanas
    1.17
    刳
    1.16
     farb
    1.15
    Activations Density 0.000%

    No Known Activations