© 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. 16935
    Prev
    Next
    INDEX
    Explanations

    blind spots or tests

    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
    esan
    -0.09
    ä¸įçĦ¶
    -0.09
    ehler
    -0.09
     обÑĢазованиÑı
    -0.09
    ally
    -0.08
    ¤
    -0.08
    utive
    -0.08
    инÑĭ
    -0.08
    à¸ĵ
    -0.08
     Alam
    -0.08
    POSITIVE LOGITS
    fold
    0.10
    è¡Įéķ¿
    0.09
    ï¸
    0.09
    оÑģÑĤи
    0.09
    OperationException
    0.09
    鸳
    0.09
    éĹ¾
    0.09
    otope
    0.09
    vertise
    0.09
    ously
    0.09
    Activations Density 0.009%

    No Known Activations