© 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. 22-GEMMASCOPE-RES-131K
    7. 129050
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on capitalized tokens—proper names, acronyms, and other named-entity–style words.

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

    EG poisoning; Lagos event; Winter War

    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
     and
    -1.07
     between
    -0.99
    crania
    -0.94
     caminar
    -0.91
     twij
    -0.90
     realizó
    -0.89
    遴
    -0.89
     Шев
    -0.88
     tomó
    -0.86
    3
    -0.85
    POSITIVE LOGITS
     daha
    1.26
    🖕
    1.12
    formin
    1.10
     whereas
    1.07
     murale
    1.02
    boten
    0.97
     perez
    0.97
    tempts
    0.95
    vaila
    0.94
    odles
    0.93
    Activations Density 0.143%

    No Known Activations