© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 6964
    Prev
    Next
    INDEX
    Explanations

    numbers followed by units or years

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

    This neuron activates on ordinary running-text words in the main body of a scientific paper (i.e. the continuous‐prose sentences), as opposed to headings, dates, or metadata.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ālā
    -0.77
    DIFF
    -0.77
     Impf
    -0.75
     секунд
    -0.70
    ートル
    -0.70
    がち
    -0.70
    移除
    -0.69
     moment
    -0.69
     verdades
    -0.69
    SFER
    -0.68
    POSITIVE LOGITS
     integrated
    1.16
    Integrated
    1.02
     Integrated
    0.93
    integrated
    0.91
    jumu
    0.91
     fb
    0.87
     Islamabad
    0.83
     dataset
    0.83
    fb
    0.82
     luminosity
    0.82
    Activations Density 0.030%

    No Known Activations