© 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. 22-GEMMASCOPE-RES-131K
    4. 79804
    Prev
    Next
    INDEX
    Explanations

    medical citations

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

    citation references in academic papers marked with [@].

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

    The neuron detects isolated “B” tokens used as reference or citation labels.

    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
    <bos>
    -1.73
     if
    -1.59
     in
    -1.58
    を
    -1.40
    は
    -1.27
     
    -1.26
     dienst
    -1.19
    ։
    -1.18
     reizen
    -1.17
    を受け
    -1.14
    POSITIVE LOGITS
     wonderful
    1.77
    凄く
    1.70
    塍
    1.70
     tremendous
    1.68
    為に
    1.66
    際には
    1.64
    ほとん
    1.63
    埝
    1.62
     ними
    1.62
     captivated
    1.61
    Activations Density 0.063%

    No Known Activations