© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 83696
    Prev
    Next
    INDEX
    Explanations

    questions and executing code

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

    The neuron fires on instruction/action keywords—typically verbs (often in infinitive or imperative form) and section labels (like ISSUE or PUBLISH) that signal steps or commands in technical instructions.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    𐬹
    -3.44
    翃
    -3.28
    鶫
    -3.16
    𓁹
    -3.00
    
    -2.95
     trois
    -2.83
    苤
    -2.80
    鶿
    -2.77
    粜
    -2.77
    ౧
    -2.75
    POSITIVE LOGITS
    e
    2.86
    </td>
    2.84
     plethora
    2.67
    沇
    2.61
     Almost
    2.47
     Giving
    2.47
     Adding
    2.41
     Getting
    2.39
     Also
    2.39
     Perhaps
    2.38
    Activations Density 0.000%

    No Known Activations