© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 84554
    Prev
    Next
    INDEX
    Explanations

    received; accepted dates

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

    The neuron specifically picks out the numeric elements (digits and related month names) in the “Received” and “Accepted” date lines of a paper’s 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_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
    But
    -1.80
     denselben
    -1.69
    
    -1.65
    ”,
    -1.63
    m
    -1.59
     gröss
    -1.54
    嵎
    -1.52
    ”،
    -1.50
    涑
    -1.50
    Its
    -1.47
    POSITIVE LOGITS
    Barbados
    1.35
    !
    1.33
    Brasil
    1.31
    México
    1.31
    Jörg
    1.30
     kiş
    1.30
    Tó
    1.29
    Invoices
    1.27
    interopRequire
    1.24
    Monique
    1.23
    Activations Density 0.118%

    No Known Activations