© 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. 3940
    Prev
    Next
    INDEX
    Explanations

    increased risk or negative outcome

    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
    ظاÙħ
    -0.09
    oriously
    -0.09
    anything
    -0.08
    cip
    -0.08
    ä¸įçĶ¨æĭħå¿ĥ
    -0.08
    Flush
    -0.08
    uilt
    -0.08
     Yus
    -0.08
    æĹłè¾ľ
    -0.08
    éĵ¢
    -0.08
    POSITIVE LOGITS
    åĽ°æĥij
    0.10
    ä¸įç¡®å®ļæĢ§
    0.10
    绢
    0.09
    失ä¸ļ
    0.09
    æĹ·
    0.09
    ntag
    0.09
     воÑģ
    0.09
    çIJ¨
    0.08
    çĦ¦èĻij
    0.08
    /de
    0.08
    Activations Density 0.208%

    No Known Activations