© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 48433
    Prev
    Next
    INDEX
    Explanations

    when September follows a comma

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    errupt
    -0.10
    á»Ļi
    -0.09
     Stark
    -0.09
    volent
    -0.09
    è±¹
    -0.09
    iculty
    -0.09
    acters
    -0.09
    裳
    -0.08
    imum
    -0.08
    çĴ
    -0.08
    POSITIVE LOGITS
    ally
    0.11
    ively
    0.11
    imes
    0.10
    iles
    0.09
    -most
    0.08
    yan
    0.08
    åĢĻ
    0.08
    طا
    0.08
    lund
    0.08
    ared
    0.08
    Activations Density 0.046%

    No Known Activations