© 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. 48750
    Prev
    Next
    INDEX
    Explanations

    Dwyer, Duane, Dwight

    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
    åŃIJéĩĮ
    -0.09
    ãģ»ãģĨ
    -0.09
    iner
    -0.09
    é»ĺ
    -0.09
    ä¾Ī
    -0.09
    Ãło
    -0.09
    falls
    -0.09
    æ¸ł
    -0.09
    æŃ¤åŁºç¡Ģä¸Ĭ
    -0.09
    ä¸Ģå¼Ģå§ĭå°±
    -0.09
    POSITIVE LOGITS
    à¸Ńาศ
    0.08
    .react
    0.08
    OI
    0.08
    STALL
    0.08
    CLE
    0.08
    ëĥIJ
    0.08
    _sink
    0.08
    ðŁı»
    0.08
    uter
    0.08
    .inject
    0.08
    Activations Density 0.004%

    No Known Activations