© 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. 103531
    Prev
    Next
    INDEX
    Explanations

    technical code snippets

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

    This neuron is triggered by numeric tokens—especially line‐number, version, and diff‐metadata numbers—in code and patch/diff contexts.

    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
     chère
    -0.73
     сва
    -0.71
     laisser
    -0.68
    versies
    -0.68
    ANTAGES
    -0.68
     मिला
    -0.66
     elimin
    -0.65
     externas
    -0.65
    Vtbl
    -0.65
    unku
    -0.65
    POSITIVE LOGITS
     Integrate
    0.68
    Para
    0.65
     erwartet
    0.64
    woł
    0.63
     stärker
    0.63
    ゲン
    0.61
    ствует
    0.61
     jim
    0.61
    ampilkan
    0.60
    solicit
    0.60
    Activations Density 0.098%

    No Known Activations