© 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. 22802
    Prev
    Next
    INDEX
    Explanations

    is followed by common phrases or negations

    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
    åĦĴ
    -0.09
    kle
    -0.08
    æªĢ
    -0.08
    满æĢĢ
    -0.08
    eps
    -0.08
    gee
    -0.08
    责任ç¼ĸè¾ij
    -0.08
    使çĶ¨æĿĥ
    -0.08
     ÑģÑĤо
    -0.08
    Ī
    -0.08
    POSITIVE LOGITS
    æĦ§
    0.11
    éĹ®é¢ĺæĺ¯
    0.10
    æ³¾
    0.10
     always
    0.09
    åĨįå¤ļ
    0.09
    两款
    0.09
    _EL
    0.09
    éļ°
    0.08
    gend
    0.08
    æ¦Ĥ念
    0.08
    Activations Density 0.068%

    No Known Activations