© 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. 23911
    Prev
    Next
    INDEX
    Explanations

    board members or actions

    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
    gence
    -0.10
    èĭįçĻ½
    -0.09
    çģ«çĤ¬
    -0.09
    ibri
    -0.09
     hardcore
    -0.09
     Hed
    -0.09
    åħļå§Ķ
    -0.09
    asley
    -0.09
     pov
    -0.09
     wes
    -0.09
    POSITIVE LOGITS
    æĪIJåijĺ
    0.11
    å¾Ĺä½ı
    0.11
    anvas
    0.09
    ware
    0.08
     MEMBER
    0.08
    uncan
    0.08
    (Board
    0.08
    å¨ĺ
    0.08
    chers
    0.08
    _member
    0.08
    Activations Density 0.015%

    No Known Activations