© 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. 37955
    Prev
    Next
    INDEX
    Explanations

    surprise and confusion

    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
    å¾ħ
    -0.10
    pending
    -0.08
     pending
    -0.08
    éķ¿æľŁä»¥æĿ¥
    -0.08
    è¿Ħä»Ĭ为æŃ¢
    -0.08
    åIJ¸åħ¥
    -0.08
    Pending
    -0.08
    æĿIJæĸĻ
    -0.08
    gay
    -0.08
    çİ°ä»»
    -0.08
    POSITIVE LOGITS
    æĬ±æĢ¨
    0.14
    åĩºçİ°éĹ®é¢ĺ
    0.12
    éĺĥ
    0.11
    æĢ¨
    0.11
    åĩºçİ°
    0.10
    æĮĩè´£
    0.10
    åĵĢ
    0.10
    åĩºçı¾
    0.09
    æĬķè¯ī
    0.09
    chaft
    0.09
    Activations Density 0.035%

    No Known Activations