© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 129656
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on tokens that are parts of personal proper names (i.e. named-entity person components).

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

    names following prefixes

    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
    為に
    -2.58
     and
    -2.39
    つづく
    -2.27
    を行います
    -2.22
    ところに
    -2.14
    をします
    -2.06
    にします
    -2.02
    おわりに
    -2.02
    をしました
    -1.98
    を行い
    -1.96
    POSITIVE LOGITS
    などが
    2.30
    <bos>
    2.30
    があり
    2.23
    がありますが
    2.14
    がある
    1.95
    などを
    1.93
    も
    1.90
    を
    1.88
    などは
    1.87
     herido
    1.87
    Activations Density 0.130%

    No Known Activations