© 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. 38945
    Prev
    Next
    INDEX
    Explanations

    but, not, always, sometimes

    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¸Ģç»ı
    -0.09
    çļĦä¼łç»Ł
    -0.09
    achs
    -0.09
    çħ§æł·
    -0.09
    æĬ¬
    -0.08
    æ©«
    -0.08
    å¯Įæ°ij
    -0.08
    横
    -0.08
    æģ¢å¤įæŃ£å¸¸
    -0.08
    åĵªæĢķ
    -0.08
    POSITIVE LOGITS
    å°ijæķ°
    0.11
    å¤ļæķ°
    0.10
    æł¹æį®ä¸įåIJĮ
    0.10
    çī¹åĪ«æĺ¯åľ¨
    0.09
    åĪĹ举
    0.09
    erton
    0.09
     nét
    0.09
    ä¹Łæľīä¸ĢäºĽ
    0.09
    .times
    0.08
    REP
    0.08
    Activations Density 0.138%

    No Known Activations