© 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. 44550
    Prev
    Next
    INDEX
    Explanations

    width and height

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

    The neuron activates on numeric tokens—especially measurements or parameters like widths, heights, scales, and other dimension values.

    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
     its
    -1.28
     and
    -1.24
    -...
    -1.21
    --,
    -1.18
     мужские
    -1.17
    =[],
    -1.04
    theres
    -1.04
     kila
    -1.03
     женские
    -1.02
    llog
    -1.02
    POSITIVE LOGITS
    ADVANTAGES
    1.26
    美味しそう
    1.15
     действительно
    1.12
     иногда
    1.09
     از
    1.09
     from
    1.09
    Regardless
    1.08
    Vaata
    1.05
     também
    1.04
     должен
    1.03
    Activations Density 0.045%

    No Known Activations