© 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. 73767
    Prev
    Next
    INDEX
    Explanations

    WH followed by acronyms or proper nouns

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

    The neuron selectively activates on uppercase “WH” acronyms or tokens (i.e. tokens beginning with “WH”).

    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
    魠
    -2.94
    
    -2.91
     drenched
    -2.69
     consentimento
    -2.61
     steeped
    -2.55
     adorned
    -2.44
    
    -2.44
     freaking
    -2.44
    ............
    -2.38
     anún
    -2.38
    POSITIVE LOGITS
    ꦁ
    2.58
    ing
    2.55
    s
    2.53
    ة
    2.48
    d
    2.38
    <i>
    2.36
     is
    2.28
    WH
    2.25
    and
    2.11
    at
    2.09
    Activations Density 0.002%

    No Known Activations