© 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. 11991
    Prev
    Next
    INDEX
    Explanations

    alternating or alternate

    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
    voor
    -0.09
     Wholesale
    -0.08
    uet
    -0.08
    èµ°å¾Ĺ
    -0.08
     nasty
    -0.08
    ì²ĺ
    -0.08
    cape
    -0.08
    æĭĶ
    -0.07
    croft
    -0.07
    erton
    -0.07
    POSITIVE LOGITS
    ately
    0.12
    éķ³
    0.12
    羣åģĩ
    0.10
    äº¤æĽ¿
    0.10
     ме
    0.10
    vens
    0.09
    enate
    0.09
     altern
    0.09
    ingly
    0.09
    ently
    0.09
    Activations Density 0.027%

    No Known Activations