© 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. 9879
    Prev
    Next
    INDEX
    Explanations

    weight topics

    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
    ../../../
    -0.11
    ä¸įäºĨ
    -0.10
    ickness
    -0.10
    atics
    -0.10
    æĹºçĽĽ
    -0.10
    wich
    -0.09
    工信
    -0.09
    èĨĽ
    -0.09
    姹
    -0.08
    å¾ij
    -0.08
    POSITIVE LOGITS
    lifting
    0.12
    lessness
    0.12
    less
    0.11
    /power
    0.10
    ings
    0.10
    åľºåĿĩ
    0.10
     mẽ
    0.10
    çłĿ
    0.10
    lessly
    0.10
    ened
    0.10
    Activations Density 0.046%

    No Known Activations