© 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. 55596
    Prev
    Next
    INDEX
    Explanations

    quantities and importance

    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
    æľĢåĸľæ¬¢çļĦ
    -0.10
    æ¹¾åĮº
    -0.10
    omed
    -0.09
    çħ§
    -0.09
    æĴ¬
    -0.09
    æľĢ主è¦ģçļĦ
    -0.09
    å¿ĥ仪çļĦ
    -0.09
    å±¥
    -0.09
    ä¼į
    -0.09
    é¹Ħ
    -0.09
    POSITIVE LOGITS
     nature
    0.13
    6
    0.11
    2
    0.11
    3
    0.11
    5
    0.11
    7
    0.11
    0
    0.11
    8
    0.11
    1
    0.10
    nature
    0.10
    Activations Density 0.247%

    No Known Activations