© 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. 26041
    Prev
    Next
    INDEX
    Explanations

    confident, confessional, confusion, confederation, confirms

    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
    -0.13
    fully
    -0.09
     thiá»ĥu
    -0.09
    lot
    -0.09
    itudes
    -0.09
    atically
    -0.09
    unde
    -0.08
    ograd
    -0.08
    ices
    -0.08
    tings
    -0.08
    POSITIVE LOGITS
    licting
    0.13
    usted
    0.11
    licted
    0.10
    ined
    0.09
    anst
    0.09
    iona
    0.09
    licts
    0.09
    erto
    0.09
    /Instruction
    0.09
    idences
    0.09
    Activations Density 0.063%

    No Known Activations