© 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. 56754
    Prev
    Next
    INDEX
    Explanations

    output followed by specific words

    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
    ãģıãĤĭ
    -0.10
    注æĺİæĿ¥æºIJ
    -0.09
    ernet
    -0.09
    urer
    -0.09
    mint
    -0.09
    ides
    -0.09
    upp
    -0.09
    åıijå±ķæľºéģĩ
    -0.09
    fur
    -0.09
    çªģåĩºéĹ®é¢ĺ
    -0.08
    POSITIVE LOGITS
    ting
    0.18
    ted
    0.14
    =output
    0.13
    wards
    0.12
    TING
    0.12
    ters
    0.12
     Ñĩего
    0.10
    (Output
    0.10
    StrLn
    0.09
    éĿŀåĩ¡
    0.09
    Activations Density 0.097%

    No Known Activations