© 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. 48828
    Prev
    Next
    INDEX
    Explanations

    extra solution, extra tabs, extra talent, extra Credits, extra column, extra edge

    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
    æIJĢ
    -0.11
    gMaps
    -0.09
    soever
    -0.09
    åħ¨å¤©
    -0.09
    allee
    -0.09
    epad
    -0.09
    /her
    -0.09
    plotlib
    -0.09
    elf
    -0.09
    夹
    -0.09
    POSITIVE LOGITS
    ordinary
    0.16
    CTION
    0.14
    (extra
    0.11
    Bonus
    0.10
    ocular
    0.10
    çķĻæĦı
    0.10
    -extra
    0.09
     Bonus
    0.09
    bonus
    0.09
    ordin
    0.09
    Activations Density 0.035%

    No Known Activations