© 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. 29796
    Prev
    Next
    INDEX
    Explanations

    sought to, expressed concern, drawing the, With MBTA, ably predicted, day after, becoming the, launched a, eted up, pushed back

    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
    3
    -0.09
    5
    -0.09
    2
    -0.09
    1
    -0.09
    4
    -0.08
    为æĪij们
    -0.08
    8
    -0.08
    éªļ
    -0.08
    _pan
    -0.08
    6
    -0.08
    POSITIVE LOGITS
    è¯Ńæ°Ķ
    0.09
     Billion
    0.09
    æĺ¨æĹ¥
    0.09
    æĦŁåıĹåĪ°äºĨ
    0.09
     ÐĴаÑĪ
    0.09
    /LICENSE
    0.09
    ç¦Ĭ
    0.09
    ä¹Łåľ¨
    0.09
    æĢı
    0.09
    uden
    0.09
    Activations Density 0.169%

    No Known Activations