© 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. 48958
    Prev
    Next
    INDEX
    Explanations

    flirt, skirt, filibuster, ship, rituals, display, forskolin, Fitbit

    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
    rise
    -0.10
    eps
    -0.09
    lund
    -0.09
    ubs
    -0.09
     trilogy
    -0.08
    çĻ¸
    -0.08
    LAG
    -0.08
    nock
    -0.08
    esian
    -0.08
    é»İæĺİ
    -0.08
    POSITIVE LOGITS
    avery
    0.10
    ation
    0.10
     nouve
    0.09
    ../../../
    0.09
     fancy
    0.09
     Hazard
    0.09
    念头
    0.09
    è°ij
    0.08
    ingly
    0.08
    åĵ¨
    0.08
    Activations Density 0.002%

    No Known Activations