© 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. 50413
    Prev
    Next
    INDEX
    Explanations

    repeated word again

    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
    çĪ¶
    -0.08
     lap
    -0.07
     oÄį
    -0.07
    æ§IJ
    -0.06
    cro
    -0.06
    inate
    -0.06
     gy
    -0.06
    omar
    -0.06
     Male
    -0.06
    å¯Ĩå¯Ĩ
    -0.06
    POSITIVE LOGITS
    again
    0.18
     Again
    0.17
     again
    0.16
    Again
    0.15
     tampoco
    0.13
    åıĪæĺ¯ä¸Ģ
    0.12
     nuevamente
    0.12
     AGAIN
    0.11
    åıĪæĺ¯ä¸Ģ个
    0.11
    ãģĵãģ¡ãĤīãĤĤ
    0.11
    Activations Density 0.028%

    No Known Activations