© 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. 101080
    Prev
    Next
    INDEX
    Explanations

    legal filings and specific technical tasks

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

    The neuron activates primarily on Japanese grammatical particles—especially the object‐marker “を.”

    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
     quindi
    -2.31
    fordshire
    -2.25
    /"
    -2.23
     and
    -2.17
    ]-->
    -2.06
    byes
    -1.99
    >-->
    -1.99
    ("
    -1.89
    ness
    -1.88
    ogues
    -1.87
    POSITIVE LOGITS
     Another
    2.22
     Versuche
    1.84
     into
    1.73
    Another
    1.73
     Gründe
    1.72
     same
    1.68
     cláss
    1.63
     another
    1.60
     Bildern
    1.60
     entirety
    1.59
    Activations Density 0.005%

    No Known Activations