© 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. 56999
    Prev
    Next
    INDEX
    Explanations

    break from

    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
     Garland
    -0.11
    éĩijåŃĹ
    -0.10
    rov
    -0.10
    inand
    -0.10
     galer
    -0.10
    coal
    -0.09
    agogue
    -0.09
    æĦıæĥ³ä¸įåĪ°
    -0.09
    å»Ĭ
    -0.09
    /gpl
    -0.09
    POSITIVE LOGITS
    ä¹ı
    0.09
    ä¹ĭéģĵ
    0.08
     reality
    0.08
    éĢļè¡Į
    0.08
    åģ¶å°Ķ
    0.08
    ble
    0.08
    å±ħå§Ķä¼ļ
    0.08
    æĹĸ
    0.08
     usual
    0.08
    å¦Ĥå®ŀ
    0.08
    Activations Density 0.010%

    No Known Activations