© 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-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 72994
    Prev
    Next
    INDEX
    Explanations

    late

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

    The neuron activates on mentions of being “late” (tardiness)—especially phrases like “late to the party,” “slightly late,” or “late for…”

    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
     angeschlossen
    -1.66
    béco
    -1.63
    uiente
    -1.62
    ztes
    -1.60
    前来
    -1.60
     emocionante
    -1.57
    怍
    -1.56
    𖥧
    -1.52
     mittler
    -1.51
    ፫
    -1.49
    POSITIVE LOGITS
     jalur
    1.44
     interessi
    1.42
    '
    1.40
    our
    1.40
     justiça
    1.40
     晚
    1.39
    萏
    1.38
     Vordergrund
    1.38
    央视
    1.38
    也是
    1.38
    Activations Density 0.040%

    No Known Activations