© 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. 8558
    Prev
    Next
    INDEX
    Explanations

    Marcel, recess

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

    The neuron consistently activates on numeric tokens or subwords containing digits (e.g. measurements, years, IDs).

    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
    ῞
    -2.81
     всѣ
    -2.59
    旰
    -2.58
    泐
    -2.52
    </strong>
    -2.52
    嶇
    -2.50
    慊
    -2.48
    雱
    -2.48
    🫃
    -2.47
    餿
    -2.45
    POSITIVE LOGITS
    3
    3.44
    h
    3.38
    k
    3.14
    to
    3.05
    w
    3.03
     Когда
    2.98
    b
    2.86
     После
    2.83
    ch
    2.83
    x
    2.81
    Activations Density 0.003%

    No Known Activations