© 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. 129833
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on sequences of capitalized personal names and initials in “author:” lists of academic paper metadata.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    authors and affiliations

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     if
    -1.82
     then
    -1.70
     now
    -1.58
     as
    -1.53
     or
    -1.52
     one
    -1.52
     on
    -1.43
     from
    -1.40
     had
    -1.40
     in
    -1.38
    POSITIVE LOGITS
     faciliter
    1.56
     participer
    1.55
     applau
    1.54
     testemun
    1.44
    Segue
    1.43
     fizer
    1.43
     frequentemente
    1.42
     humeur
    1.40
     supplé
    1.39
     accompagner
    1.39
    Activations Density 0.004%

    No Known Activations