© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 8623
    Prev
    Next
    INDEX
    Explanations

    opinion by

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

    The neuron identifies high-level section headers or separators (e.g. “OPINION OF THE COURT,” “-----Original Message-----,” “FW:,” “RE:,” 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
    úly
    -1.44
    了许多
    -1.41
     zrobić
    -1.37
    StreetMap
    -1.33
    ('/
    -1.26
    겟
    -1.26
     znaleźć
    -1.25
     wybrać
    -1.25
     kupić
    -1.24
    chläge
    -1.23
    POSITIVE LOGITS
     her
    1.45
    okkal
    1.44
    qtt
    1.42
    ganggu
    1.41
     quinze
    1.36
    tiens
    1.36
    Muffin
    1.34
     estabiliz
    1.34
    alimentos
    1.31
     of
    1.30
    Activations Density 0.011%

    No Known Activations