© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 98933
    Prev
    Next
    INDEX
    Explanations

    Doughboy Hollow, Buckeye Advocate, ALEPH Collaboration

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

    This neuron fires on the second element of compound proper names or titles (e.g. the “ling” in “sanderling,” “boom” in “jibboom,” “eye” in “buckeye,” 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
     verir
    -1.63
     urbanas
    -1.36
     ürünleri
    -1.34
     kesim
    -1.28
    Kanada
    -1.27
     měla
    -1.26
     which
    -1.22
    注意事
    -1.21
     Dasar
    -1.20
     aéreas
    -1.20
    POSITIVE LOGITS
     his
    1.56
    </em>
    1.54
     of
    1.53
    )
    1.52
     then
    1.51
     last
    1.46
     Both
    1.45
     außerdem
    1.43
     also
    1.31
    ];
    1.29
    Activations Density 2.700%

    No Known Activations