© 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. 111041
    Prev
    Next
    INDEX
    Explanations

    an online pharmacy, this invention, chemical composition

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

    the beginning of new paragraphs or sentences, often activating at periods followed by capital letters.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron is keyed to sentence boundaries, firing on period tokens and the words that start new sentences (e.g. “The,” “I,” “This,” “An,” “It,” numbers, etc.).

    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
     first
    -2.13
     increase
    -1.80
    Surprisingly
    -1.79
    When
    -1.77
    Three
    -1.69
    Perhaps
    -1.68
    Essentially
    -1.66
    Quite
    -1.65
     when
    -1.64
     before
    -1.63
    POSITIVE LOGITS
    <bos>
    1.66
    -
    
    1.60
     Dinas
    1.55
    
    1.52
    桫
    1.49
    摀
    1.48
    
    1.48
    eſ
    1.48
     goede
    1.48
     korte
    1.48
    Activations Density 0.002%

    No Known Activations