© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 6966
    Prev
    Next
    INDEX
    Explanations

    profanity and strong opinions

    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
     hubby
    -0.09
    åijµåijµ
    -0.09
    ä¸Ģç¢Ĺ
    -0.08
    ä¿ļ
    -0.08
    æĥ³ä¸įåĪ°
    -0.08
    oops
    -0.08
    çļĦèĥĮåIJİ
    -0.08
    好象
    -0.08
    亲æĥħ
    -0.08
    åıĸè¯ģ
    -0.08
    POSITIVE LOGITS
     fucking
    0.12
     fuck
    0.11
     fucks
    0.10
     shit
    0.10
    æŀģåħ¶
    0.10
     FUCK
    0.10
    åį§
    0.09
     Fucking
    0.09
    æŀģ大
    0.09
    æ½Ń
    0.08
    Activations Density 0.591%

    No Known Activations