© 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. 63049
    Prev
    Next
    INDEX
    Explanations

    equivalent, sufficiently, directly, lower, approx

    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
    åĩĭ
    -0.10
    æĢ¥åī§
    -0.10
    çļĭ
    -0.10
    WAR
    -0.09
     Ain
    -0.09
    åĭŁ
    -0.09
    .rmi
    -0.08
    è¹´
    -0.08
    å¥
    -0.08
     вн
    -0.08
    POSITIVE LOGITS
    éĴĹ
    0.10
    å°ij许
    0.09
     twice
    0.09
    æĥħåķĨ
    0.09
    åĨł
    0.09
    åĩłåĪĨ
    0.08
    é«ĺåĩº
    0.08
    èģĮæĿĥ
    0.08
     worlds
    0.08
    éĥ¨
    0.08
    Activations Density 0.039%

    No Known Activations