© 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. 37675
    Prev
    Next
    INDEX
    Explanations

    downward spiral or decline

    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
    çģ«çĤ¬
    -0.09
    .pk
    -0.09
    ayo
    -0.09
     Surge
    -0.08
    instein
    -0.08
    dek
    -0.08
    razy
    -0.08
    ç«ŁçĦ¶
    -0.08
    æĿ¡
    -0.08
    .sourceforge
    -0.08
    POSITIVE LOGITS
    ä¸ĢæŃ¥æŃ¥
    0.15
    è´¥
    0.14
    ä¸ĢæŃ¥ä¸ĢæŃ¥
    0.13
    æģ¶åĮĸ
    0.12
    丧失
    0.12
    åłķ
    0.11
    æ¸IJ
    0.11
     descent
    0.11
    丢失
    0.11
     decline
    0.10
    Activations Density 0.099%

    No Known Activations