© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 26275
    Prev
    Next
    INDEX
    Explanations

    Clear BSD, Canvas, objects, Eye, Continental

    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
    Negative Logits
    丧失
    -0.08
    ноÑģÑĤ
    -0.08
    å¿«æĬ¥
    -0.08
    æļ¨
    -0.08
    éĶ¦
    -0.08
     WithEvents
    -0.08
    å¦Ħ
    -0.08
    æĪIJçĨŁ
    -0.08
     parach
    -0.08
    balls
    -0.08
    POSITIVE LOGITS
    ät
    0.09
    anca
    0.09
    è¦ģåģļ好
    0.09
     catch
    0.09
     cabin
    0.09
     Notebook
    0.09
     ÐĵÑĢ
    0.09
    ESÄ°
    0.08
     DPR
    0.08
    tif
    0.08
    Activations Density 0.004%

    No Known Activations