© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 9263
    Prev
    Next
    INDEX
    Explanations

    negative numbers

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

    The neuron fires on the “−” sign (i.e. the minus/hyphen token, especially as used in negative numbers or subtraction).

    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
    s
    -3.73
    Not
    -3.52
    After
    -3.02
    There
    -2.92
    In
    -2.91
    On
    -2.89
    Before
    -2.89
    Even
    -2.88
    Just
    -2.86
    When
    -2.78
    POSITIVE LOGITS
    萏
    3.06
     dát
    2.98
    鶫
    2.63
    齶
    2.58
    ೢ
    2.56
    脔
    2.56
    媃
    2.52
    狨
    2.47
    ఽ
    2.45
    膑
    2.44
    Activations Density 0.001%

    No Known Activations