© 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. 58999
    Prev
    Next
    INDEX
    Explanations

    strong argument could be made

    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
    linger
    -0.10
    åģ¶
    -0.09
    åĴĮå°ļ
    -0.09
    heit
    -0.09
    ront
    -0.09
    иÑĢ
    -0.09
    æĪijåĸľæ¬¢
    -0.09
     Duncan
    -0.09
    设计åĻ¨
    -0.09
    ooter
    -0.08
    POSITIVE LOGITS
    åı¯ä»¥è¯´
    0.14
    åı¯ä»¥è¯´æĺ¯
    0.13
     argument
    0.13
    认为
    0.13
     argue
    0.12
     argued
    0.11
    argument
    0.11
    arguments
    0.11
     arg
    0.11
    .argument
    0.11
    Activations Density 0.041%

    No Known Activations