© 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
    2. Gemma-3-27B-IT
    3. 53-GEMMASCOPE-2-RES-262K
    4. 49497
    Prev
    Next
    INDEX
    Explanations

    album, figure, Avril, canzone, serieHere's the breakdown:* `MAX_ACTIVATING_TOKENS` contains 'album', 'figure', 'Avril', 'canzone', and other potential entities or parts of words.* `TOKENS_AFTER_MAX_ACTIVATING_TOKEN` shows that after 'figure' comes 'chiave' (key), after 'canzone' comes 'rap', and 'serie' appears after '/'.* `TOP_ACTIVATING_TEXTS` confirms the presence of 'album' (Lucio Battisti albums), 'figure' (figure chiave/key figures), 'Avril' (Avril 14th), and 'canzone rap' (rap song). It also contains 'Serie A'.The common theme is identifying specific items or concepts, often presented in lists or as titles/names. The identified tokens like 'album', 'figure', 'Avril', 'canzone', and 'serie' are distinct entities or categories.The phrase "album, figure, Avril, canzone, serie" directly lists these key distinct items identified by the neuron across the examples. It's concise and captures the essence of what the neuron seems to be activated by. album, figure, Avril, canzone, serie

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_53_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     свое
    0.46
    सुद्धा
    0.43
     pais
    0.43
     सन्
    0.43
     ограничен
    0.40
     GOOD
    0.40
     пред
    0.39
    дыду
    0.39
     त
    0.39
     Raceway
    0.39
    POSITIVE LOGITS
    abilità
    0.50
     servizi
    0.46
    erà
    0.44
    nému
    0.44
    ůli
    0.44
     Україні
    0.43
    importanza
    0.42
    ipy
    0.42
     забезпечення
    0.42
     rinnov
    0.42
    Activations Density 0.001%

    No Known Activations