© 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. 49980
    Prev
    Next
    INDEX
    Explanations

    nonlinear scaling behavior

    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
    èĪ¬
    -0.10
    ult
    -0.10
     med
    -0.10
    als
    -0.09
    bsp
    -0.09
    tam
    -0.08
    èĪ¬çļĦ
    -0.08
    ä¸Ģçķª
    -0.08
    -0.08
    biz
    -0.08
    POSITIVE LOGITS
    ProcessEvent
    0.09
    åĮºåĨħ
    0.09
    ÑĢÑĥб
    0.09
    ench
    0.09
    ä¸ĢåĪĩéĥ½æĺ¯
    0.08
    _echo
    0.08
    avra
    0.08
    à¸Ńม
    0.08
    æĢ§å¼º
    0.08
    åŁİå¸Ĥ建设
    0.08
    Activations Density 1.898%

    No Known Activations