© 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. 34-GEMMASCOPE-RES-131K
    7. 45388
    Prev
    Next
    INDEX
    Explanations

    potential

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

    The neuron fires strongly on the minus‐sign token (and related numeric/operator tokens), so it’s detecting arithmetic negation signs.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     though
    -1.15
     although
    -1.12
     Although
    -1.11
     specific
    -1.10
     Though
    -1.07
    Though
    -1.07
     While
    -1.06
    Although
    -1.02
     will
    -1.02
     through
    -1.02
    POSITIVE LOGITS
    "){
    1.03
    wechsel
    1.02
     muziek
    1.02
    pfung
    1.01
    ()+"
    1.01
    骉
    1.00
    ítő
    0.98
    диод
    0.96
    棡
    0.96
    />
    0.95
    Activations Density 0.003%

    No Known Activations