© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 8576
    Prev
    Next
    INDEX
    Explanations

    informal "man" interjections

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

    The neuron detects informal vocative interjections or conversational address terms (e.g. “man,” “bro,” “mate,” “maaaan”).

    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
    unakan
    -1.55
    klare
    -1.43
    읊
    -1.39
     uchun
    -1.35
    Jueves
    -1.34
    misen
    -1.34
     anhänger
    -1.31
    huriyet
    -1.31
     bygger
    -1.30
    躂
    -1.30
    POSITIVE LOGITS
     are
    1.63
     лишь
    1.52
     wirk
    1.43
    ]?.
    1.38
     deiner
    1.36
    蓠
    1.35
    #,
    1.34
     комнату
    1.34
    はこの
    1.33
     jedynie
    1.32
    Activations Density 0.006%

    No Known Activations