© 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. 33516
    Prev
    Next
    INDEX
    Explanations

    coding prefixes

    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
    人æµģ
    -0.10
    éĥģ
    -0.09
    çIJ¥
    -0.09
    æĪ¢
    -0.08
    ä¸ĢåĪĢ
    -0.08
    å¾ģæ±Ĥ
    -0.08
    嵩
    -0.08
    é»ĺ
    -0.08
    æŀī
    -0.08
    æľ¬ç½ij
    -0.08
    POSITIVE LOGITS
    åĢį
    0.10
    家åħ·
    0.10
     furniture
    0.10
    urniture
    0.10
    ä¾
    0.10
    egie
    0.09
     improv
    0.09
    身为
    0.09
    ActivityResult
    0.09
    åĤ¢
    0.09
    Activations Density 0.048%

    No Known Activations