© 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. 24117
    Prev
    Next
    INDEX
    Explanations

    features and their characteristics

    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
    ìį¨
    -0.10
    sWith
    -0.10
    yna
    -0.09
    entifier
    -0.09
    aji
    -0.09
    åįģä¸ĩ
    -0.09
    ä¸ĢåĽŀäºĭ
    -0.09
    aly
    -0.08
    竺
    -0.08
    roller
    -0.08
    POSITIVE LOGITS
     prominently
    0.13
    tte
    0.13
    amp
    0.10
    lli
    0.10
    ground
    0.10
    ally
    0.09
    лин
    0.09
    ial
    0.09
    åIJĦå¼Ĥ
    0.08
    -rich
    0.08
    Activations Density 0.048%

    No Known Activations