© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 15995
    Prev
    Next
    INDEX
    Explanations

    the token "time"

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

    This neuron responds to rare or “out‐of‐vocabulary” tokens—e.g. special‐character punctuation (en‐dashes), non‐Latin script (Greek letters), and uncommon all‐caps or proper‐name tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    stitches
    0.89
    monkey
    0.88
    滀
    0.87
    markt
    0.85
    యత్
    0.85
    𝗽
    0.84
    ゴー
    0.83
     symplectic
    0.83
    shader
    0.83
     illusory
    0.82
    POSITIVE LOGITS
    ان
    0.73
    따
    0.71
    ك
    0.70
    Méd
    0.70
    ال
    0.68
    Opts
    0.68
    BBC
    0.64
    Media
    0.63
    et
    0.63
    ể
    0.62
    Activations Density 0.000%

    No Known Activations