© 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. 18283
    Prev
    Next
    INDEX
    Explanations

    "sum" followed by "s" or "Gr"

    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
    çģ¯çģ«
    -0.10
    ograd
    -0.10
    æĿµ
    -0.09
     Cristiano
    -0.09
    sea
    -0.09
    éĴĵé±¼
    -0.09
    COOKIE
    -0.09
    OLT
    -0.09
    LING
    -0.09
    æĶ¶è´¹ç«Ļ
    -0.09
    POSITIVE LOGITS
    orda
    0.10
    aroo
    0.09
    zin
    0.09
    ç¬ij
    0.08
     Chew
    0.08
    amma
    0.08
    çĪŃ
    0.08
    unte
    0.07
    inent
    0.07
    Pocket
    0.07
    Activations Density 0.003%

    No Known Activations