© 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. 60791
    Prev
    Next
    INDEX
    Explanations

    spellings and spells

    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
    apos
    -0.10
    iliar
    -0.09
    ially
    -0.09
    åķ·
    -0.09
    æ±Łæ¹ĸ
    -0.09
    çĤľ
    -0.09
    бав
    -0.09
    Ñģка
    -0.08
    imal
    -0.08
    aily
    -0.08
    POSITIVE LOGITS
    ings
    0.13
    makers
    0.10
    itude
    0.10
    pell
    0.10
    ر
    0.10
    making
    0.09
     thoại
    0.09
    usters
    0.09
    rons
    0.09
    worm
    0.09
    Activations Density 0.007%

    No Known Activations