© 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. 2015
    Prev
    Next
    INDEX
    Explanations

    purpose of

    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
    ä¸ĢæŶ
    -0.10
    æĹ¶ä»£çļĦ
    -0.08
    å¼¢
    -0.08
     Milo
    -0.08
    ä¸Ģ代
    -0.08
    osaic
    -0.08
    ä¹¡åľŁ
    -0.08
    ä¸īæľŁ
    -0.08
    èµ·æŃ¥
    -0.08
    èĪ
    -0.08
    POSITIVE LOGITS
    fully
    0.13
    缮çļĦ
    0.13
     purpose
    0.11
    çļĦ缮çļĦ
    0.11
     Purpose
    0.11
    purpose
    0.11
    entes
    0.10
    æŨ
    0.10
    åľ¨äºİ
    0.10
    Purpose
    0.10
    Activations Density 0.065%

    No Known Activations