© 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. 128064
    Prev
    Next
    INDEX
    Explanations

    stars, pulsars, accretion

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

    The neuron is strongly activated by tokens containing numerals (e.g. scientific measurements, identifiers, or alphanumeric codes).

    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
    ⬈
    -2.52
     responsabilité
    -2.16
    culously
    -2.13
    潆
    -2.09
     oiseau
    -2.06
    一点点
    -2.05
    mtliche
    -2.03
     ſou
    -1.98
    HASAN
    -1.95
    蘩
    -1.95
    POSITIVE LOGITS
    B
    2.56
    H
    2.25
    R
    2.23
    A
    2.23
    U
    2.19
    S
    2.14
    ↵↵
    2.14
    G
    2.11
            
    2.08
     ‘
    1.99
    Activations Density 0.008%

    No Known Activations