© 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. 10-GEMMASCOPE-RES-131K
    7. 90632
    Prev
    Next
    INDEX
    Explanations

    connection string, timeout, link, televised

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

    The neuron lights up on isolated content tokens naming specific technical or proper‐noun concepts—things like event names, API function names, or specialized technical terms.

    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
     fulfilling
    -2.56
    ти
    -2.53
     categorically
    -2.44
     
    -2.42
     importantly
    -2.39
     comprising
    -2.33
    がなく
    -2.33
     preferably
    -2.33
    ТЫ
    -2.27
     scrutin
    -2.20
    POSITIVE LOGITS
    полне
    2.75
     ними
    2.59
    ть
    2.48
    咔嚓
    2.47
     solucionar
    2.42
     aclarar
    2.38
     goof
    2.34
     TAMBIÉN
    2.33
     gustado
    2.31
     paździer
    2.30
    Activations Density 0.003%

    No Known Activations