© 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. 38633
    Prev
    Next
    INDEX
    Explanations

    opinion shall not constitute precedent

    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
    åĪ«è¯´
    -0.11
    ãģ¨ãģĦãģ£ãģ¦
    -0.10
    æ¿Ĵ临
    -0.10
    童年
    -0.09
    å®ī康
    -0.09
    ãģıãĤĭ
    -0.09
    æ·¼
    -0.09
    ı
    -0.09
    æľīä¸Ģ天
    -0.08
    -ever
    -0.08
    POSITIVE LOGITS
    aire
    0.11
    aires
    0.10
    POSITE
    0.09
     Watkins
    0.09
    kest
    0.09
    nels
    0.09
    ราà¸Ħา
    0.09
    <<<
    0.08
    ure
    0.08
    (Void
    0.08
    Activations Density 0.032%

    No Known Activations