© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100736
    Prev
    Next
    INDEX
    Explanations

    non followed by words

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

    The neuron activates on occurrences of the negative prefix “non-” (i.e. the “non” in hyphenated negation 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
    vereignty
    -1.73
     of
    -1.67
                
    -1.66
    ou
    -1.63
    thole
    -1.61
    栊
    -1.57
             
    -1.55
    ul
    -1.55
    IN
    -1.48
          
    -1.48
    POSITIVE LOGITS
     EVERY
    1.68
    ).
    1.48
     what
    1.38
     or
    1.37
     any
    1.31
     gối
    1.31
     mesto
    1.29
     motivadoras
    1.28
     Evolutionary
    1.28
    jalanan
    1.27
    Activations Density 0.046%

    No Known Activations