© 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. 12073
    Prev
    Next
    INDEX
    Explanations

    direction and orientation

    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
    åħ±è¯Ĩ
    -0.10
    æºIJ头
    -0.10
    éĴ¿
    -0.10
    archives
    -0.09
    pler
    -0.09
     McConnell
    -0.09
    åĪ¬
    -0.09
    æĸ½
    -0.09
    arez
    -0.09
    è¡¿
    -0.09
    POSITIVE LOGITS
    æĸ¹ä½į
    0.15
     direction
    0.13
    æĸ¹åIJij
    0.13
    direction
    0.12
     Direction
    0.12
     directions
    0.12
    举åįĹ
    0.12
    (Direction
    0.11
     Directions
    0.11
    çļĦæĸ¹åIJij
    0.11
    Activations Density 0.040%

    No Known Activations