© 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. 10-GEMMASCOPE-RES-131K
    7. 44872
    Prev
    Next
    INDEX
    Explanations

    filled, laden, soaked, drenched

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

    This neuron detects descriptive compound adjectives—especially hyphenated or concatenated forms ending in “-strewn,” “-laden,” “-filled,” “-sodden,” “-infested,” etc.

    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
     such
    -1.79
    </sub>
    -1.70
     is
    -1.67
     has
    -1.66
    cayaan
    -1.60
     -
    -1.60
    さま
    -1.53
     Cedex
    -1.52
     this
    -1.52
     f
    -1.49
    POSITIVE LOGITS
    푉
    1.83
    Teknik
    1.80
     übri
    1.78
     ſever
    1.76
     biß
    1.75
    
    1.70
    Spes
    1.65
     świąte
    1.63
     COULD
    1.63
     obrazy
    1.58
    Activations Density 0.034%

    No Known Activations