© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 43193
    Prev
    Next
    INDEX
    Explanations

    foods and sounds

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

    The main thing this neuron does is spot onomatopoeic “pop,” “puff,” or “crackle” terms (often in food names like popcorn, puffed rice, nachos, cotton candy, or in popping/crackling sounds).

    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
    .”
    -2.72
    ⚋
    -2.44
    ↵
    -2.23
    The
    -2.20
    进一步
    -2.19
    It
    -2.11
    of
    -2.08
    在他的
    -2.02
     Siehe
    -2.02
    这些
    -2.02
    POSITIVE LOGITS
    凩
    2.50
     erst
    2.30
    赑
    2.03
     ehemal
    2.02
    ୪
    2.02
     altes
    2.00
     máme
    1.97
    鹛
    1.94
     junger
    1.86
    sandalia
    1.84
    Activations Density 0.003%

    No Known Activations