© 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. 35328
    Prev
    Next
    INDEX
    Explanations

    specific instances after "this"

    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
    æĮ½
    -0.10
     onboard
    -0.09
    è¦ģçľĭ
    -0.09
     onsite
    -0.09
    gtk
    -0.09
    ä¼łæİĪ
    -0.09
     Saud
    -0.09
    .flash
    -0.09
    è¤Ĭ
    -0.09
    ä¼łè¾ĵ
    -0.09
    POSITIVE LOGITS
     behalf
    0.25
     occasion
    0.13
    /off
    0.12
    ymous
    0.11
     occasions
    0.10
    athan
    0.10
    -site
    0.10
    eday
    0.10
    -campus
    0.09
    etime
    0.09
    Activations Density 0.043%

    No Known Activations