© 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. Qwen3.5-9B Base
    5. 15-QWENSCOPE-RES-64K
    7. 33676
    Prev
    Next
    INDEX
    Explanations

    erroneously / improperly / falsely / wrongly

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    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
    è®¤çľŁéĺħ读
    -0.08
    unexpected
    -0.07
    éģĩåĪ°çļĦéĹ®é¢ĺ
    -0.07
    éģĩåĪ°çļĦ
    -0.07
     reasonably
    -0.07
    ushman
    -0.07
    æľīçĤ¹åĦ¿
    -0.07
    ä¸įå¤ļ
    -0.07
    lems
    -0.07
    ansom
    -0.07
    POSITIVE LOGITS
    æĪĸä¸į
    0.13
    åĴĮä¸į
    0.11
    /false
    0.11
    /un
    0.10
    éĶĻ误çļĦ
    0.09
    licherweise
    0.09
    erweise
    0.09
     incorrect
    0.09
    æĪĸæľª
    0.09
    /non
    0.08
    Activations Density 0.115%

    No Known Activations