© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 4396
    Prev
    Next
    INDEX
    Explanations

    pronouns like he, him, his

    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
    ä»ĸä»¬åľ¨
    -0.11
    ä»ĸ们
    -0.11
    éĥ½å¾Ī
    -0.10
    åIJĦèĩª
    -0.10
     themselves
    -0.10
    ä»ĸ们çļĦ
    -0.09
     ÙĦدÙĬ
    -0.09
    娣
    -0.09
    leur
    -0.08
     Ñģами
    -0.08
    POSITIVE LOGITS
    /her
    0.24
    åĴĮä»ĸçļĦ
    0.23
     himself
    0.21
    /she
    0.20
    èĢģ人家
    0.19
     Himself
    0.14
    ä¿©
    0.14
    ather
    0.12
    aped
    0.12
    .her
    0.12
    Activations Density 0.662%

    No Known Activations