© 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. 72883
    Prev
    Next
    INDEX
    Explanations

    people's names

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

    The neuron fires strongly on short, capitalized name fragments (e.g. “Li,” “He,” “Po,” “De,” “Ba,” etc.), i.e. pieces of proper nouns in people’s or publication titles.

    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
    軼
    -2.64
    _
    -2.63
    焢
    -2.31
    I
    -2.22
    鈊
    -2.13
     using
    -2.06
    .
    -2.03
    If
    -2.03
    -
    -2.03
    结语
    -2.02
    POSITIVE LOGITS
    (!__
    2.55
     '
    2.39
    饧
    2.34
     was
    2.14
    哺乳
    2.06
    ⮜
    2.06
    百合
    2.00
    回路
    1.95
    </em>
    1.94
    ፁ
    1.94
    Activations Density 0.238%

    No Known Activations