© 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. 44000
    Prev
    Next
    INDEX
    Explanations

    place names

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

    The neuron activates strongly on proper nouns that are geographic place names (towns, villages, rivers, etc.).

    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
     how
    -1.66
     like
    -1.62
    </b>
    -1.30
     통해
    -1.27
     why
    -1.26
    讓我們
    -1.26
    </u>
    -1.22
     обуви
    -1.20
     simply
    -1.19
     joão
    -1.16
    POSITIVE LOGITS
     to
    1.76
    abinieri
    1.39
    ]")]
    1.34
     damaligen
    1.33
    "…
    1.30
    两
    1.30
    ;(
    1.27
    !!,
    1.27
    :...
    1.27
     two
    1.27
    Activations Density 0.570%

    No Known Activations