© 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. 34-GEMMASCOPE-RES-131K
    4. 58989
    Prev
    Next
    INDEX
    Explanations

    section introduction citations

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

    This neuron specifically detects the closing curly brace that ends a section-identifier tag (e.g., “}” in “{#sec1-sensors-2002641}”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     enough
    -0.93
     šo
    -0.85
     kaikk
    -0.85
    Meskipun
    -0.83
     pirm
    -0.82
     although
    -0.81
     miljö
    -0.81
     vienas
    -0.81
     dabar
    -0.80
    のですね
    -0.80
    POSITIVE LOGITS
     thậm
    0.84
    CAI
    0.81
     Eds
    0.81
     WIR
    0.79
     bahkan
    0.78
     Loon
    0.78
     Steak
    0.77
     francis
    0.77
    ApiModelProperty
    0.76
     восто
    0.76
    Activations Density 0.004%

    No Known Activations