© 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. 38274
    Prev
    Next
    INDEX
    Explanations

    said in a

    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
    åīĮ
    -0.10
    åĴĸ
    -0.10
    ç¯ĩ
    -0.09
    顽强
    -0.09
    eger
    -0.09
    彩èī²
    -0.08
    è¯ķè¯ķ
    -0.08
    åĪº
    -0.08
    ä¸Ńå¿ĥåŁİåĮº
    -0.08
    èij³
    -0.08
    POSITIVE LOGITS
    .apps
    0.09
    ä½ľåĿĬ
    0.08
    çľĦ
    0.08
     bloom
    0.08
    ÑĨиÑĤ
    0.08
    人åı£
    0.08
    ayo
    0.08
    [](
    0.08
    ãħ£
    0.08
     игÑĢ
    0.08
    Activations Density 0.026%

    No Known Activations