© 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. 4096
    Prev
    Next
    INDEX
    Explanations

    scattered points or plot

    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
    acity
    -0.09
    soever
    -0.09
    itud
    -0.09
    ặ
    -0.08
    éĶIJ
    -0.08
    adecimal
    -0.08
    ãģĵãģ¨
    -0.08
    avior
    -0.08
     Äijỡ
    -0.08
    arem
    -0.08
    POSITIVE LOGITS
    å¼ĢæĿ¥
    0.12
    ingly
    0.10
    tee
    0.10
    ä¼Ĭæĭī
    0.10
    edReader
    0.10
    loff
    0.10
    æķ£
    0.09
    PHONE
    0.09
    edException
    0.09
    dü
    0.09
    Activations Density 0.023%

    No Known Activations