© 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. 22-GEMMASCOPE-RES-131K
    7. 15254
    Prev
    Next
    INDEX
    Explanations

    nodes, organization, site, names, associated

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

    The neuron fires on prominent technical nouns or “section” terms (like article, node, pane, environment, topic, component) that serve as headings or key document‐structure labels.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    だったら
    -1.34
     our
    -1.34
     didn
    -1.30
     Knochen
    -1.28
     Lovely
    -1.28
     Hauses
    -1.25
    phism
    -1.23
    火的
    -1.20
    nonumber
    -1.20
     уже
    -1.20
    POSITIVE LOGITS
     can
    1.58
     eſ
    1.55
    ſo
    1.54
     hidupnya
    1.32
    There
    1.32
    】:
    1.30
     ſer
    1.30
     itulah
    1.30
     discí
    1.30
     Gaetano
    1.29
    Activations Density 0.186%

    No Known Activations