© 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-3-27B-IT
    5. 27-GEMMASCOPE-2-TRANSCODER-262K
    7. 199563
    Prev
    Next
    INDEX
    Explanations

    I am unable to provide an explanation based on the provided lists. The `MAX_ACTIVATING_TOKENS` list contains only punctuation and special characters ('.', '\`'), and the `TOKENS_AFTER_MAX_ACTIVATING_TOKEN` list is empty. This lack of sequential text or clear linguistic markers makes it impossible to identify a recognizable pattern or concept that governs the neuron's behavior.If you can provide lists with more substantial text tokens, I can attempt to identify the pattern

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_27_width_262k_l0_small_affine
    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
    可爱
    0.42
     photoelectron
    0.41
     goddesses
    0.40
    ক্ষিতে
    0.38
     correspondents
    0.38
     breezes
    0.38
    हमारी
    0.38
     heathen
    0.38
    くださ
    0.38
    ര്‍ത്ത
    0.37
    POSITIVE LOGITS
     Lue
    0.41
     Max
    0.39
     Lith
    0.38
     Rory
    0.38
     Carr
    0.37
     Manu
    0.37
     det
    0.37
     several
    0.37
     Rie
    0.37
    all
    0.37
    Activations Density 0.001%

    No Known Activations