© 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. 22-GEMMASCOPE-RES-131K
    7. 35257
    Prev
    Next
    INDEX
    Explanations

    late or belated events/occurrences

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

    The neuron detects words or phrases that denote lateness or belated timing (e.g. “late,” “belated,” “retrospective”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     first
    -1.05
    discus
    -0.92
     пър
    -0.91
    galeria
    -0.90
    先は
    -0.90
    ouvrir
    -0.90
    describing
    -0.88
    いま
    -0.87
    ۔۔
    -0.87
    idées
    -0.86
    POSITIVE LOGITS
    late
    1.15
     Late
    1.06
     добав
    1.05
     addition
    1.05
     late
    1.04
     amends
    0.98
    Late
    0.95
     parezca
    0.94
     позд
    0.94
    setVisibility
    0.91
    Activations Density 0.068%

    No Known Activations