© 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. 113710
    Prev
    Next
    INDEX
    Explanations

    holding or held

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

    This neuron fires on occurrences of the verb “hold” (including its inflected forms held, holding).

    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
    .
    -2.78
     Another
    -2.19
     There
    -1.97
    をも
    -1.94
     Earlier
    -1.88
    Another
    -1.85
     Surprisingly
    -1.85
     about
    -1.85
     Его
    -1.81
     Segen
    -1.80
    POSITIVE LOGITS
    </strong>
    2.61
    ”)
    2.25
    )
    2.20
    ”:
    2.19
    ⸂
    2.16
    涑
    2.14
    ”;
    2.13
    UAWEI
    2.11
    ).
    2.09
    </i>
    2.08
    Activations Density 0.038%

    No Known Activations