© 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. 57846
    Prev
    Next
    INDEX
    Explanations

    sugar

    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
    æīĭä¸ŃçļĦ
    -0.10
     Douglas
    -0.10
    éĢ¼
    -0.08
    诽
    -0.08
     burgers
    -0.08
    æķ°åĢ¼
    -0.08
     поÑĤеÑĢÑı
    -0.08
    åĭŁéĽĨ
    -0.08
    alli
    -0.08
    OrCreate
    -0.08
    POSITIVE LOGITS
    èĶĹ
    0.12
    ç³ĸ
    0.12
    çĶľ
    0.11
    èľľ
    0.11
    çĶľèľľ
    0.11
    éħIJ
    0.10
     sugar
    0.10
     Sugar
    0.09
    sweet
    0.09
     Caval
    0.09
    Activations Density 0.029%

    No Known Activations