© 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. 101688
    Prev
    Next
    INDEX
    Explanations

    coding specific terms

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

    words or tokens that are part of technical or specialized terminology, particularly in non-English languages (like Russian or Chinese characters) and programming/technical contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    This neuron detects tokens that end with common inflectional or derivational English suffixes (e.g. –ed, –ing, –less, –tion, –ular).

    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
     only
    -1.93
     not
    -1.44
    2
    -1.40
     no
    -1.40
    Only
    -1.38
     goles
    -1.30
     first
    -1.28
    N
    -1.28
     same
    -1.28
    F
    -1.25
    POSITIVE LOGITS
     such
    1.80
    ють
    1.52
    唖
    1.48
     cereales
    1.45
     though
    1.45
     especias
    1.43
    lakukan
    1.38
    zijn
    1.36
     являются
    1.35
    ganet
    1.34
    Activations Density 0.212%

    No Known Activations