© 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. 49975
    Prev
    Next
    INDEX
    Explanations

    sentinel / surveillance / guard

    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
    æ²½
    -0.10
    und
    -0.09
    меÑĢ
    -0.09
    uard
    -0.09
    å¾Ĺèµ·
    -0.09
    ãĥĭãĥ³ãĤ°
    -0.09
    ied
    -0.09
    æĪIJ为ä¸Ģ个
    -0.09
    ãģĹãģ¾ãģĦ
    -0.09
    oder
    -0.08
    POSITIVE LOGITS
    ä¹ĭä½ľ
    0.10
    ä¹ĭ家
    0.10
    IZED
    0.09
    aneously
    0.09
    ties
    0.09
    enced
    0.09
    ted
    0.09
     isempty
    0.09
    æ³ķ人
    0.09
    åIJĦæł·
    0.08
    Activations Density 0.002%

    No Known Activations