© 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. 30027
    Prev
    Next
    INDEX
    Explanations

    conjunctions indicating consequence or transition

    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
    ä¸įå¾Ĺå·²
    -0.11
    arih
    -0.10
    orem
    -0.09
    ç»Ļä½łä»¬
    -0.09
     allerdings
    -0.09
    æ¯Ķå¦Ĥ说
    -0.09
    æĬ¤
    -0.09
    æİīäºĨ
    -0.09
    çĦ¶å¾Į
    -0.09
    ands
    -0.09
    POSITIVE LOGITS
    è¶Ĭæĺ¯
    0.09
     there
    0.09
     they
    0.09
     it
    0.09
    reek
    0.09
     Cabinets
    0.08
     (?)
    0.08
     Gar
    0.08
    è¶Ĭ
    0.08
    éļıå¤Ħ
    0.08
    Activations Density 0.135%

    No Known Activations