© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 38243
    Prev
    Next
    INDEX
    Explanations

    help but

    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
    IRT
    -0.11
    cert
    -0.09
    èѬ
    -0.09
    åĿļå®ļä¸į移
    -0.09
    æĪij说
    -0.09
    è°ģçŁ¥éģĵ
    -0.09
    ä¸įå°ij人
    -0.08
    ä¸ļçķĮ
    -0.08
    ä¸į容æĺĵ
    -0.08
    æ¸ħç®Ĺ
    -0.08
    POSITIVE LOGITS
    ably
    0.11
    ä½ı
    0.10
    odyn
    0.09
    emente
    0.09
    edin
    0.09
    ently
    0.09
    üb
    0.09
    为ä¼ģä¸ļ
    0.09
    çĦ¶æĺ¯
    0.08
    being
    0.08
    Activations Density 0.003%

    No Known Activations