© 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. 45026
    Prev
    Next
    INDEX
    Explanations

    onboard, backstage, onstage performer, heroine

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

    The neuron fires on compound adverbial/location words like “onboard” and “backstage” that indicate being in or at a particular place.

    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
    ’
    -3.14
    as
    -2.72
    }
    -2.64
    ?
    -2.55
     '
    -2.44
    /
    -2.44
    '
    -2.42
    ]
    -2.36
    \\
    -2.27
    ,”
    -2.16
    POSITIVE LOGITS
    嵙
    2.72
     但是
    2.55
    割と
    2.48
    𖣘
    2.47
     Есть
    2.44
    籏
    2.44
    他的
    2.42
    獴
    2.39
     Так
    2.38
     Многие
    2.36
    Activations Density 0.004%

    No Known Activations