© 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. 17483
    Prev
    Next
    INDEX
    Explanations

    cogna, cogload, cogged

    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
    uw
    -0.09
    èĢģåħµ
    -0.09
    ınd
    -0.09
    fter
    -0.09
    æݾ
    -0.09
    lide
    -0.08
    æĸ¹åIJij
    -0.08
    htar
    -0.08
    é¹ħ
    -0.08
    å¼łè´´
    -0.08
    POSITIVE LOGITS
    stone
    0.09
    çļĦç»ıåİĨ
    0.09
    ÐľÐŀ
    0.09
     ØŃÙĤ
    0.09
    zilla
    0.09
    ä¸įæ¸ħ
    0.09
    ellation
    0.08
    ä½ķæŶ
    0.08
    hole
    0.08
    éĴ©
    0.08
    Activations Density 0.010%

    No Known Activations