© 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. 62606
    Prev
    Next
    INDEX
    Explanations

    many an / much religion before noun

    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
    çĭĻ
    -0.10
    aldi
    -0.09
    /logging
    -0.09
    ingo
    -0.09
    ç»Ļ大家
    -0.09
    ãģıãĤĵ
    -0.08
    éĥ¨åĪĨåľ°åĮº
    -0.08
    鸱
    -0.08
     pige
    -0.08
    è¿Ļä¸ĢåĪĩ
    -0.08
    POSITIVE LOGITS
    ä¸įçĦ¶
    0.12
    /all
    0.09
    éĥĬ
    0.09
     Shen
    0.08
    gal
    0.08
    ɵ
    0.08
    rlen
    0.08
    abb
    0.08
    ãĥ¼ãĥ«
    0.08
    rzy
    0.08
    Activations Density 0.004%

    No Known Activations