© 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. 56459
    Prev
    Next
    INDEX
    Explanations

    in turn

    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
    æ°ij
    -0.10
    å¤Ħæĸ¹
    -0.08
    (up
    -0.08
    æĦ¿æĻ¯
    -0.08
    çĥĪ
    -0.08
    este
    -0.08
    度åģĩ
    -0.08
    ech
    -0.08
     Carson
    -0.08
     Cookbook
    -0.08
    POSITIVE LOGITS
    åıįè¿ĩæĿ¥
    0.16
    æŃ¤åŁºç¡Ģä¸Ĭ
    0.11
     turn
    0.11
     далÑĮн
    0.11
     itself
    0.10
    è¿ĽèĢĮ
    0.10
     Ñģобой
    0.10
    bles
    0.10
    ä»ĸåıĪ
    0.09
    åıĪ被
    0.09
    Activations Density 0.031%

    No Known Activations