© 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. 57132
    Prev
    Next
    INDEX
    Explanations

    handling 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
    åİ»æİī
    -0.10
    ä¼ļéķ¿
    -0.09
    lke
    -0.09
    ndon
    -0.08
    stab
    -0.08
    åį´åıijçİ°
    -0.08
    åĨĢ
    -0.08
    chen
    -0.08
    æĪŁ
    -0.08
    建ç«ĭèµ·
    -0.08
    POSITIVE LOGITS
     handling
    0.14
     Handling
    0.11
    handling
    0.11
    Handling
    0.11
    å°¤åħ¶æĺ¯åľ¨
    0.11
    åĩłæ¬¡
    0.10
    ä¸įåIJĪ
    0.10
    对å¾ħ
    0.10
    æ¶īå«Į
    0.10
    äºī
    0.10
    Activations Density 0.085%

    No Known Activations