© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 10442
    Prev
    Next
    INDEX
    Explanations

    while working which why the

    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
     Chennai
    -0.12
     Tamil
    -0.10
    èĮĦ
    -0.10
    å©Ĩ
    -0.09
     Kerala
    -0.09
     Vish
    -0.09
     Bangalore
    -0.09
     Spy
    -0.09
     Vive
    -0.09
    åį°åº¦
    -0.09
    POSITIVE LOGITS
    å·´åŁºæĸ¯åĿ¦
    0.11
    inger
    0.10
    ç¼ł
    0.09
    Pakistan
    0.09
     shedding
    0.09
    å·¥åķĨ
    0.08
    ç®Ģä»ĭ
    0.08
    éĺ¿å¯Įæ±Ĺ
    0.08
    æĶ»åĿļ
    0.08
     Afghanistan
    0.08
    Activations Density 0.180%

    No Known Activations