© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25501
    Prev
    Next
    INDEX
    Explanations

    code tokens and data retrieval

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

    The neuron is detecting programming‐related tokens—identifiers and keywords you’d see in code (e.g. loop directives like “repeat”/“in,” collection types like “List”/“Page,” variable names like “data,” etc.).

    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
     antworten
    -0.83
    Ứng
    -0.82
     on
    -0.82
    createCanvas
    -0.82
    начала
    -0.82
    上げます
    -0.81
     Category
    -0.81
    初
    -0.81
     installed
    -0.81
     División
    -0.80
    POSITIVE LOGITS
     pietre
    0.98
    🐩
    0.97
     secoli
    0.96
    ִּ
    0.94
     Alamy
    0.94
    سمبر
    0.93
    Библи
    0.91
    Личная
    0.91
    cabul
    0.91
    Materiale
    0.90
    Activations Density 0.003%

    No Known Activations