© 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. 22336
    Prev
    Next
    INDEX
    Explanations

    performance and efficiency

    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
    屯
    -0.10
    ç®ķ
    -0.10
    HEL
    -0.10
    OrUpdate
    -0.09
    ResourceManager
    -0.09
    ikon
    -0.09
    çł
    -0.09
    ç®Ń
    -0.09
    å®ĿçŁ³
    -0.08
    ç«ĭ项
    -0.08
    POSITIVE LOGITS
     performance
    0.17
    performance
    0.16
    æĢ§èĥ½
    0.15
    表çİ°
    0.15
     performances
    0.14
     Performance
    0.13
    Performance
    0.13
    ä¸įä½³
    0.13
    formance
    0.13
     PERFORMANCE
    0.13
    Activations Density 0.092%

    No Known Activations