© 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. 49224
    Prev
    Next
    INDEX
    Explanations

    un- prefix

    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
    æŀ¸
    -0.10
    ä¸įç®Ĺ
    -0.10
    æıIJ款
    -0.09
    æŁĶ软
    -0.09
    å¿ĥ缮ä¸Ń
    -0.09
    罡
    -0.09
    èĢģ头
    -0.09
    æĹłéĻIJ
    -0.08
    çī¹åĪ«æĺ¯åľ¨
    -0.08
    èĭŀ
    -0.08
    POSITIVE LOGITS
     intact
    0.13
     untouched
    0.12
    没æľī太大
    0.10
    ä¸įåıĹ
    0.10
    没
    0.09
    ä¸įåģļ
    0.09
    åıĹåĪ°
    0.09
     unchanged
    0.09
     hy
    0.09
    åĩı
    0.09
    Activations Density 0.074%

    No Known Activations