© 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. 17687
    Prev
    Next
    INDEX
    Explanations

    end of phrase/sentence

    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
    outh
    -0.09
    ared
    -0.09
    pan
    -0.08
    æ·ij
    -0.08
    éĹ½
    -0.08
    丰ç¡ķ
    -0.08
    usi
    -0.08
    çĭ¸
    -0.08
    è¿Ļä¹Łæĺ¯
    -0.08
    ä¸ŃåĽ½åĽ½éĻħ
    -0.08
    POSITIVE LOGITS
    å¹´å¹´åºķ
    0.09
    _ke
    0.09
     suce
    0.09
    Å¡e
    0.09
    æĦ¿
    0.08
    ilde
    0.08
    祷
    0.08
    (delegate
    0.08
    å¼ĢçĿĢ
    0.08
    anford
    0.08
    Activations Density 0.004%

    No Known Activations