© 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. 23752
    Prev
    Next
    INDEX
    Explanations

    disapproval or vulgarity

    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
    IfNeeded
    -0.09
    æ½ĩæ´Ĵ
    -0.09
    èİİ
    -0.09
    yı
    -0.08
    ãĥĥãĤ·
    -0.08
    Sid
    -0.08
    nnen
    -0.08
    åıĮåıĮ
    -0.08
    海滨
    -0.08
    sons
    -0.08
    POSITIVE LOGITS
    ulent
    0.10
    uous
    0.09
    pile
    0.09
    ening
    0.09
    tern
    0.09
    kop
    0.09
    izens
    0.08
    face
    0.08
    ulous
    0.08
    ota
    0.08
    Activations Density 0.007%

    No Known Activations