© 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. 58726
    Prev
    Next
    INDEX
    Explanations

    risk of or risk being

    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
    åĿļå®ļä¸į移
    -0.11
    è¿Ļäºĭ
    -0.10
    æĮ½
    -0.09
    æĶ»åĬ¿
    -0.09
    raw
    -0.09
    äºĮåįģåĽĽ
    -0.09
    æĽ·
    -0.08
     Cres
    -0.08
    .undefined
    -0.08
    Prince
    -0.08
    POSITIVE LOGITS
    沦
    0.10
    üh
    0.09
    ä¸Ģ头
    0.09
    ä¸Ģè¾ĪåŃIJ
    0.09
     cart
    0.09
    æĪIJ为
    0.09
    ange
    0.08
    apan
    0.08
     being
    0.08
    aise
    0.08
    Activations Density 0.007%

    No Known Activations