© 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. 84191
    Prev
    Next
    INDEX
    Explanations

    code and foreign language text

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

    The neuron fires on SQL‐style relational keywords (e.g. ON, FROM) that introduce table sources or joins.

    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
    <blockquote>
    -1.02
    0
    -1.01
    9
    -1.00
    レンチ
    -0.93
    7
    -0.92
    8
    -0.92
    LoadScene
    -0.89
    4
    -0.87
    受害
    -0.86
    2
    -0.86
    POSITIVE LOGITS
    անի
    1.05
    を整
    0.98
    满是
    0.98
    ירים
    0.92
     appris
    0.90
     bocetos
    0.88
     나는
    0.88
    begre
    0.87
     questa
    0.87
     당신
    0.87
    Activations Density 0.008%

    No Known Activations