© 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. 38583
    Prev
    Next
    INDEX
    Explanations

    dates or numbers with delimiters

    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
    oth
    -0.11
    /her
    -0.10
    oner
    -0.10
    wich
    -0.10
    sing
    -0.09
    ái
    -0.09
    ä¸Ģå®ļçļĦ
    -0.09
    ewart
    -0.08
    edes
    -0.08
    pare
    -0.08
    POSITIVE LOGITS
    atile
    0.10
    åĨįåĬłä¸Ĭ
    0.09
    нÑĮ
    0.09
    untary
    0.08
     дело
    0.08
    idian
    0.08
     Fool
    0.08
    ARIABLE
    0.08
    è¿ĺæľī
    0.08
    lude
    0.08
    Activations Density 0.365%

    No Known Activations