© 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. 64722
    Prev
    Next
    INDEX
    Explanations

    code symbols or common programming punctuation

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

    The neuron fires strongly on numeric literal tokens (especially integer constants) in source code.

    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
     to
    -1.34
     berühmten
    -1.30
    
    -1.30
    Хотя
    -1.27
     whereas
    -1.26
    lawayo
    -1.26
     Whereas
    -1.22
    despite
    -1.20
     เพราะ
    -1.19
     allgemeinen
    -1.13
    POSITIVE LOGITS
     he
    1.27
     بودند
    1.13
     fühlte
    1.09
     glaubte
    1.08
     ovan
    1.08
     basé
    1.08
    ,
    1.07
     würd
    1.06
    tică
    1.05
     combinar
    1.03
    Activations Density 0.001%

    No Known Activations