© 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. 21578
    Prev
    Next
    INDEX
    Explanations

    Annals, grail, pains, Straits

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

    The neuron fires on the key content words that head multi-word named entities or fixed expressions (e.g. “Straits” in Straits Settlements, “Holy” in Holy Grail, “Annals” in Annals of Tigernach, “pains” in aches and pains).

    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
     mauva
    -3.25
    腚
    -3.17
     новом
    -3.16
    榦
    -3.03
    屺
    -3.00
    笮
    -2.95
    渙
    -2.94
    龊
    -2.91
    Eglise
    -2.91
    诳
    -2.89
    POSITIVE LOGITS
    4
    3.61
    }
    3.42
    2
    3.39
    It
    3.33
    ing
    3.31
    With
    3.31
    Of
    3.27
    7
    3.27
    All
    3.23
    This
    3.17
    Activations Density 0.035%

    No Known Activations