© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 112255
    Prev
    Next
    INDEX
    Explanations

    saber, lightsaber, Sabre

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

    The neuron is a “name‐spotter” – it strongly activates on single tokens that are proper names or technical terms (e.g. Exercise Saber, the verb/noun saber, the acronym SABRE, lightsaber, luciferase, luteum, 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
    「
    -2.45
     =
    -2.45
     '
    -2.41
    \
    -2.39
     lazos
    -2.28
    霂
    -2.28
     addition
    -2.23
    ve
    -2.23
     ihr
    -2.22
     ''
    -2.19
    POSITIVE LOGITS
     It
    2.41
    ￳
    2.39
     菇
    2.38
    Ꮌ
    2.30
    𖣘
    2.30
     セックス
    2.27
     puteți
    2.25
    VIATIONS
    2.22
    ኗ
    2.20
     您
    2.16
    Activations Density 0.005%

    No Known Activations