© 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. 10-GEMMASCOPE-RES-131K
    4. 12448
    Prev
    Next
    INDEX
    Explanations

    UI elements and data descriptors

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

    The neuron is picking out programming‐specific identifiers and technical tokens—things like class or variable names, URL/URI fragments, API terms, and other code‐related labels.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     at
    -1.34
    嗉
    -1.15
     Obtén
    -1.10
    諳
    -1.10
    Idade
    -1.10
     escuchadas
    -1.09
    見ても
    -1.06
    に行ってきました
    -1.05
    Preparazione
    -1.05
    AILED
    -1.05
    POSITIVE LOGITS
     only
    1.38
     telah
    1.16
     aforementioned
    1.09
     anecd
    1.06
     Myself
    1.05
     kalk
    1.05
    ---
    1.02
     загряз
    1.02
    只有
    1.02
    asnya
    1.01
    Activations Density 0.176%

    No Known Activations