© 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. 26340
    Prev
    Next
    INDEX
    Explanations

    current

    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
    BEL
    -0.10
    ends
    -0.10
    ðŁı»
    -0.10
    ode
    -0.10
    jax
    -0.10
    ore
    -0.10
    é¢ı
    -0.09
    ðŁı¼
    -0.09
    ending
    -0.09
    ctors
    -0.09
    POSITIVE LOGITS
    izr
    0.12
    aneous
    0.11
    vature
    0.11
    åķĨæĬ¥
    0.10
    mente
    0.10
    idad
    0.10
     Affairs
    0.10
    izacao
    0.10
    Millis
    0.10
    eer
    0.10
    Activations Density 0.099%

    No Known Activations