© 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. 57009
    Prev
    Next
    INDEX
    Explanations

    def + following words

    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
    ä¼łæĿ¥
    -0.10
    ç»Ļä½łä»¬
    -0.09
    .predicate
    -0.08
    åIJij举
    -0.08
    chem
    -0.08
    ----------------------------------------------------------------------------
    -0.08
    çĮ¿
    -0.08
    åIJ»
    -0.08
     ok
    -0.08
    åģ¶
    -0.08
    POSITIVE LOGITS
    :def
    0.11
    (def
    0.10
    缺éĻ·
    0.10
    aint
    0.10
     def
    0.09
    éĻ·
    0.09
    CUS
    0.09
    -def
    0.09
    def
    0.09
    .def
    0.09
    Activations Density 0.029%

    No Known Activations