© 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. 3189
    Prev
    Next
    INDEX
    Explanations

    rightly or for reasons

    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
    _exc
    -0.11
    æĭIJ
    -0.09
    wig
    -0.09
    çİ¢
    -0.09
    zer
    -0.08
    æłĩå¿ĹæĢ§
    -0.08
    ÙĨج
    -0.08
     lạc
    -0.08
    Fake
    -0.08
    åıªéľĢè¦ģ
    -0.08
    POSITIVE LOGITS
    没éĶĻ
    0.17
     rightly
    0.16
     пÑĢавилÑĮно
    0.15
    æŃ£ç¡®
    0.15
     correctly
    0.14
    æŃ£ç¡®çļĦ
    0.14
     Äijúng
    0.13
    æŃ£ãģĹãģĦ
    0.13
     correct
    0.13
     rightfully
    0.12
    Activations Density 0.086%

    No Known Activations