© 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. 36370
    Prev
    Next
    INDEX
    Explanations

    torture

    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
    ossier
    -0.09
    sis
    -0.09
    ë¡Ŀ
    -0.08
    LEM
    -0.08
    sst
    -0.08
     trá»Ŀi
    -0.08
    ä¸Ģåı·
    -0.08
    ness
    -0.08
    sWith
    -0.08
    éric
    -0.08
    POSITIVE LOGITS
    uchi
    0.10
    -resistant
    0.09
    иÑħ
    0.09
    uring
    0.09
    éĽĨåĽ¢èĤ¡ä»½
    0.09
    ç»ĵå±Ģ
    0.09
    пед
    0.09
    ç¼
    0.09
    åľ°éģĵ
    0.08
    亲æĥħ
    0.08
    Activations Density 0.004%

    No Known Activations