© 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. 36077
    Prev
    Next
    INDEX
    Explanations

    "one" or "single" followed by an action or unit

    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
     Ballard
    -0.10
    羣çIJĨ
    -0.09
    æĦļ
    -0.09
    enburg
    -0.09
    cede
    -0.09
    æļ¹
    -0.09
    æ¯Ķéĩį
    -0.09
    è°²
    -0.09
    inee
    -0.08
    顺çĿĢ
    -0.08
    POSITIVE LOGITS
    ä¸Ģ次æĢ§
    0.20
    ä¸Ģ次
    0.17
    ä¸Ģ举
    0.15
    ä¸ĢæŃ¥
    0.13
    ä¸Ģä¸ĭåŃIJ
    0.12
    ä¸Ģä¸ĭåŃIJå°±
    0.12
    ä¸Ģéģĵ
    0.12
    ä¸Ģç«Ļå¼ı
    0.11
    ä¸Ģåı£æ°Ķ
    0.11
    ä¸Ģé¡¿
    0.11
    Activations Density 0.073%

    No Known Activations