© 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. 61267
    Prev
    Next
    INDEX
    Explanations

    (parenthesized sources)

    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
    é¦Ĩ
    -0.11
    館
    -0.10
    ä»Ģä¹Īåij¢
    -0.10
    çļĦ带é¢Ĩ
    -0.09
    abouts
    -0.09
    _stride
    -0.09
    åºŃ
    -0.09
    çļĦéĤ£ç§į
    -0.09
    atement
    -0.08
    çīĸ
    -0.08
    POSITIVE LOGITS
    borg
    0.09
    licable
    0.08
     Cra
    0.08
    èĩ¬
    0.08
    ÙĪغ
    0.08
    oreach
    0.07
    cone
    0.07
    ä¼ļéķ¿
    0.07
     converse
    0.07
    plete
    0.07
    Activations Density 0.150%

    No Known Activations