© 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. 54143
    Prev
    Next
    INDEX
    Explanations

    negative descriptors

    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
    éħIJ
    -0.10
    æľīçĤ¹
    -0.09
    åĤ¬
    -0.09
    è°ĥçļ®
    -0.09
    æı©
    -0.09
    ä¸į太好
    -0.09
    ä¸į太
    -0.09
    æıĴ
    -0.08
    ä¸Ģè§Ĵ
    -0.08
    æĬ¹
    -0.08
    POSITIVE LOGITS
    èĪ¬çļĦ
    0.09
    èĢĮåıĪ
    0.09
    ä¸į顾
    0.09
    宪
    0.08
    mong
    0.08
    unas
    0.08
    å·´æĸ¯
    0.08
    åĩĮ
    0.08
    .POS
    0.08
    å¾·å°Ķ
    0.08
    Activations Density 0.079%

    No Known Activations